Why triethylene glycol rather than one of its neighbours?

Because it sits at the useful compromise point. A smaller glycol carries over into the treated gas more readily and is lost faster; a larger one is more viscous, harder to pump and strip, and gives a smaller gain in performance than the extra handling trouble costs. Triethylene glycol combines strong affinity for water with low enough vapour pressure to stay in the loop and with a regeneration behaviour that ordinary field equipment can achieve. It is the default for that reason, not because the others cannot absorb water.

What limits how dry the gas can be made?

The concentration of the lean glycol leaving the regenerator. Water in the treated gas approaches equilibrium with the liquid it last touched, so the driest gas the absorber can produce is fixed by how much water the regenerator managed to remove. Adding contact stages or circulating more glycol moves the result closer to that limit, but neither can pass it. When a unit will not meet its target, the regenerator is the first place to look, not the contactor.

Does raising the circulation rate improve the result?

Only up to a point, and beyond it the effect reverses. More glycol through the reboiler means less energy available per unit of liquid to strip the water out, so the lean concentration falls and the ceiling comes down with it. Higher flow also raises carryover and shortens residence time in the surge and filtration train. Circulation is the control most often pushed the wrong way during a wet-gas complaint, because the instinct to send more absorbent is stronger than the instinct to check what condition the absorbent is in.

What causes glycol losses that do not show up as a leak?

Three mechanisms account for most of it. Mechanical carryover from the contactor, when mist elimination is damaged, fouled or simply outmatched by the gas rate. Vaporisation into the treated gas, which rises as the contactor runs warmer. Losses out of the regenerator overhead, either as entrainment or because the still column is not doing its job of returning glycol to the reboiler. A loss audit that compares make-up volume against these three routes usually finds the answer faster than a search for drips.

How should incoming glycol be specified and checked?

Specify assay with the homologue distribution reported separately rather than folded into one number, water content by a stated method, acidity, chloride, ash or residue, and colour. On receipt, keep a retained sample from every lot and run at least water, acidity and colour before the drum or tanker goes into the surge vessel. The reason is trend detection: a single certificate describes a moment, whereas a series of retained samples describes a direction, and direction is what tells you whether the loop is degrading.

How TEG dehydration works: absorption and regeneration

Technical article · Eapearl Chemical · 5 August 2026

Gas dehydration by glycol absorption looks simple on a flow sheet and is unforgiving in practice. Almost every performance problem traces back to one relationship: the regenerator decides the ceiling, and the contactor only decides how close you get to it.

What the absorber is actually doing

The molecule at the centre of the process is triethylene glycol, CAS 112-27-6, formula C₆H₁₄O₄, molar mass 150.17 g/mol. It carries two hydroxyl groups at the ends and two ether oxygens along the chain, which gives it four sites capable of hydrogen bonding with water. That is the whole mechanism. Water in the gas phase finds a liquid that binds it more strongly than the hydrocarbon around it does, and partitions into that liquid.

Nothing is consumed and nothing reacts. The glycol is a carrier that picks water up where it is cool and gives it back where it is hot, and the economics of the unit rest on the fact that this cycle can be repeated many thousands of times before the carrier itself has to be replaced. Two further properties make the carrier practical: it is far less volatile than water, so it stays in the loop rather than leaving with the treated gas, and it tolerates repeated heating better than the smaller glycols do.

Equilibrium sets the ceiling, contact decides the approach

Inside the contactor, gas rises through trays or structured packing while lean glycol flows down against it. At every stage the two phases move toward equilibrium, and the water content the gas would reach at perfect equilibrium with the liquid at the top of the column is the best the unit can ever do. That equilibrium value depends on how concentrated the lean glycol is and on the conditions in the column. It does not depend on how many trays are installed.

Trays and circulation rate determine only how near the outlet gas comes to that limit. This distinction is worth stating in plain terms because it explains a common and expensive mistake. When treated gas fails a dryness specification, crews often reach first for the circulation rate, on the reasonable-sounding theory that more absorbent removes more water. If the lean glycol is weak, more of it does not help, and frequently makes matters worse.

The regenerator is the whole process in miniature

Rich glycol leaving the contactor is let down in pressure, flashed to recover entrained hydrocarbon, filtered, warmed against the returning lean stream, and fed to a still column above a reboiler. Heat drives the water off overhead; concentrated glycol collects below and returns to the top of the contactor. Every constraint on the process meets here.

The reboiler cannot simply be run hotter until the glycol is as dry as desired, because the glycol itself begins to break down thermally, and the products of that breakdown are acidic. Acidity accelerates corrosion, corrosion products foul heat transfer surfaces, fouling creates local hot spots, and hot spots accelerate degradation. This is the loop that turns a neglected unit into a failing one, and it is why acidity in the circulating inventory is a leading indicator rather than a housekeeping figure.

Raising the ceiling without cooking the inventory

Because the reboiler is bounded, field practice has developed several ways to strip more water at the same heat input. All of them work by lowering the partial pressure of water above the liquid in the regeneration section.

Stripping gas — a small dry gas stream injected into the reboiler or a packed section beneath the still, sweeping water vapour away from the liquid surface.
Vacuum regeneration — reducing pressure over the reboiler so that water leaves at a lower temperature.
Azeotropic and solvent-assisted schemes — circulating a light hydrocarbon that carries water overhead and is condensed, separated and returned.
Improved still internals — packing and reflux arrangements that return glycol to the reboiler instead of losing it overhead, which raises efficiency and cuts losses at the same time.

Each buys lean concentration at a cost in equipment, utilities or complexity. The design question is never which is best in the abstract, but how much dew point depression the contract actually requires and what the cheapest route to that number is.

Reading a wet result by symptom

Observation	What it usually points to
Lean glycol assay low, reboiler at normal duty	Circulation too high for the heat available, or still column flooding
Lean assay fine, treated gas still wet	Poor contact: tray damage, channelled packing, inlet distribution, or gas rate outside design
Make-up volume rising steadily	Carryover from the contactor, or glycol leaving the still overhead
Rising acidity in retained samples	Thermal degradation; check reboiler duty, fouling and flash separation
Foaming, level instability, erratic dryness	Liquid or solid carryover from the inlet, surfactant contamination, or hydrocarbon in the rich stream
Dark inventory, filters blinding quickly	Corrosion products and degradation residue accumulating faster than filtration removes them

Keeping the inventory clean

The inlet separator matters more than any control setting. Produced water, brine, compressor lubricant, corrosion inhibitor and fine solids all arrive with the gas, and none of them leave with the water vapour overhead. They accumulate in the circulating inventory, where brine leaves chloride, lubricant and inhibitor promote foaming, and solids abrade pumps and blind filters. A particulate filter followed by an activated carbon stage handles the ordinary burden; nothing downstream compensates for a separator that is undersized or bypassed.

Foaming deserves separate mention because it damages performance in two directions at once. Foam reduces effective contact in the column, so the gas leaves wetter, and it carries liquid overhead, so glycol is lost. Treating the symptom with antifoam while ignoring what reached the contactor buys time and nothing else.

What the purchase specification has to carry

Order by identifier rather than by trade name, and ask for the homologue distribution to be reported rather than summed, since diethylene glycol and tetraethylene glycol both behave differently in the loop from the nominal product and both appear as normal minor components. Ask for water, acidity, chloride, residue and colour, and agree the analytical methods in writing, because two laboratories reporting different methods will disagree about the same drum. Note also that ethylene glycol appears in gas facilities for a quite different duty, hydrate inhibition upstream, and mixing the two inventories through shared hoses is a recurring and avoidable cause of off-specification results.

Specifications, lot certificates and retained-sample arrangements for glycols used in field service are discussed against a named duty through our contact page.