
Article Content
A triethylene glycol dehydrator is built to remove water. It also removes a share of the H2S, and it concentrates what it removes into a small regenerator overhead. That overhead is often treated as a minor vent. It is not. On a sour gas plant or a battery dehydrating solution gas, the still-column vent, the flash tank, and the glycol filters are where dissolved sulfide leaves the gas and becomes an exposure, a corrosion, and an emissions problem.
Most operators know the gas is sour. Fewer have a number on the glycol itself, or a plan for the regenerator vent that is sized on that number rather than on a sweet-gas assumption.
How H2S Gets Into the Glycol
TEG contacts the gas in the contactor. H2S is soluble in the rich glycol, along with CO2, hydrocarbons, and water. The amount that dissolves depends on the partial pressure of H2S in the gas, the temperature, the glycol circulation rate, and how clean the glycol is. A higher H2S partial pressure and a colder contactor put more sulfide into the rich stream.
The rich glycol then goes to the flash tank, if there is one, and to the regenerator. In the flash tank, light hydrocarbons and some of the dissolved acid gas break out. In the regenerator, heat drives off water and most of the remaining H2S and CO2. The still-column overhead is therefore a small gas stream that can be far richer in H2S, on a concentration basis, than the main gas was.
A second path is contamination. Sour produced water in the inlet separator carryover, iron sulfide from upstream corrosion, and sulfate-reducing bacteria in a poorly maintained glycol system all add sulfide that was not in the original gas analysis. A dehydrator that was fine on sweet gas can sour after a well change or after a long shutdown with wet glycol left in the system.
Where the Problem Shows Up
The still-column vent is the main release. On many skids it is a short stack or a pipe to a safe location. If that pipe ends at working height, or if it is tied into a vent header that is not scrubbed, the regenerator becomes a continuous low-volume sour source. The concentration can be high enough to matter even when the mass rate is small.
The flash tank is the second point. Gas off the flash tank is often routed to fuel or to a flare. If it is routed to fuel without a check, it can put H2S into the fuel header. If it is flared without a scrubber, it is a sulfur emission. Either way it needs a number.
Filters and the glycol itself are the third. Iron sulfide and other solids load the filters. A filter change on a sour glycol system is an exposure job, not a routine mechanical job. Rich glycol sampled at the contactor outlet, and lean glycol sampled after the regenerator, tell you whether the sulfide is being stripped or whether it is accumulating.
Corrosion follows the wet acid gas. The regenerator, the still column, the reboiler vapor space, and any carbon-steel piping that sees rich glycol or regenerator overhead are the usual locations. Pitting under deposits and general wastage in the still column are common once the gas has been sour for a while and the metallurgy was chosen for sweet service.
What to Measure
Three numbers sort a dehydrator. H2S in the inlet gas, H2S in the rich glycol or in the flash-tank gas, and H2S in the still-column overhead. The inlet gas tells you the load. The overhead tells you what the vent or the scrubber has to handle. The glycol number tells you whether sulfide is accumulating in the loop.
Sample the overhead at the still column, not only at a distant vent stack where air has diluted it. Sample the glycol from a closed point and run it before the sulfide strips. A gas-only tube at the contactor outlet does not describe the regenerator.
| Sample point | What it tells you | When to take it |
| Contactor inlet gas | Load entering the system | After any well or feed change |
| Flash-tank gas | What is going to fuel or flare | During normal circulation |
| Still-column overhead | What the vent or scrubber must handle | At operating reboiler temperature |
| Rich and lean glycol | Whether sulfide is accumulating | On a set schedule, and after upsets |
Treating the Overhead
The still-column gas is the right place to treat. The volume is small and the H2S has already been concentrated. A small caustic, oxidant, or triazine scrubber, or a solid adsorbent bed, can take that stream to a safe vent or to a flare. Size it on the overhead rate at full glycol circulation and on the high inlet H2S case, not on the sweet-gas nameplate.
Routing the overhead into an existing amine regenerator vent or into a plant flare is fine if that destination is already rated and permitted for the extra sulfur. Routing it into a fuel-gas header is not, unless that header is treated and the engine or burner limit can take it. Many dehydrator vents have been tied to fuel because the original gas was sweet.
Liquid scavenger injected into the glycol can reduce the sulfide in the loop. It is a supplement, not a substitute for treating the overhead. Dose on a test of the actual rich glycol. Overdosing amine-type scavenger into TEG can add nitrogen compounds and solids, and it does not fix a vent that is already open. Keep the scavenger compatible with the glycol and with the elastomers in the pumps.
Solid beds on the overhead suit steady, low rates and sites that do not want a liquid chemical on the dehydrator skid. Protect the bed from liquid carryover. A still column that pukes glycol will foul a bed in one event. A knockout ahead of the bed is part of the design.
Corrosion and Materials
Carbon steel is common on older dehydrators and is a weak choice once the overhead is sour and wet. The still column, the vapor line, and the reboiler vapor space see the worst of it. Upgrading those sections, or accepting a higher inspection frequency and a corrosion allowance, is the mechanical half of the fix. Inhibiting the glycol helps the liquid side. It does not protect a wet sour vent line.
Filters deserve a procedure. Isolate, depressurize, and check the headspace before a housing is opened. Iron sulfide on a filter is both a solid and a source of H2S if it is acidified or wetted and agitated. Bag spent filters and keep them out of an open bin.
Bacteria and Shutdowns
Glycol systems that sit wet and stagnant can grow sulfate-reducing bacteria, especially if sulfate has entered with produced water carryover. The result is sulfide that appears after a startup even though the inlet gas has not changed. A biocide program, a clean surge tank, and not leaving rich glycol standing for weeks are the controls. Scavenger after the fact removes the sulfide. It does not stop the next batch from forming in a dead leg.
On startup after a sour history, expect the first overhead to be richer than the steady reading. Treat or route that first gas before people work around the still column.
A Dehydrator Layout That Holds on Sour Gas
For a TEG unit on variable or sour gas, the set that works is short.
- Measure inlet gas, flash-tank gas, and still-column overhead, not only the contactor outlet.
- Route the still-column vent to a scrubber, a treated flare, or a permitted vent. Do not tie it to fuel unless the fuel limit allows it.
- Knock out liquids ahead of any adsorbent on the overhead.
- Keep solids and iron sulfide out of the filters with a change procedure that includes a headspace check.
- Re-check the overhead after any feed change, and after any shutdown that left wet glycol standing.
That is a sample habit and a small vent scrubber. It is cheaper than a corroded still column and cheaper than an exposure at the regenerator.
| Option for the overhead | Best when | Watch-out |
| Small liquid scrubber | Variable H2S, chemical already on site | Level, spent chemical, carryover |
| Solid adsorbent bed | Low steady rate, remote skid | Liquid carryover, breakthrough |
| Tie to existing treated flare or vent | Destination is already rated | Permit and backpressure |
| Tie to fuel gas | Only if fuel limit and treating allow it | Engines and instruments see the H2S |
| Glycol-side scavenger only | Never as the only control | Does not fix an open vent |
Circulation Rate and the Overhead Load
Operators sometimes raise glycol circulation to chase a water dew point. That also raises the amount of H2S picked up in the contactor and the amount stripped in the regenerator. A scrubber or a bed sized on the original circulation rate can be undersized after a dew-point correction. If circulation is increased, re-check the overhead rate and the H2S concentration before assuming the vent treating still has margin.
The reverse is also true. Cutting circulation reduces the acid-gas load on the regenerator, but it does not remove sulfide already dissolved in the holdup. After a cut, the overhead can stay rich for a while as the system turns over. Sample after the change, not only before it.
Hydrocarbon pickup complicates both cases. Rich glycol that carries condensate will flash more volume in the flash tank and can foam in the regenerator. Foam carries liquid into the overhead and into any downstream scrubber or bed. An antifoam program and a clean glycol are part of H2S control on a sour dehydrator, because a foaming still column defeats a well-sized vent scrubber by sending glycol into it.
Reboiler temperature and stripping gas, if used, also move the overhead number. A hotter reboiler or a stripping-gas rate that was raised to improve lean-glycol purity will push more H2S into the still column. Note those setpoints when the overhead is sampled, and treat a setpoint change the same way as a feed change: re-measure before calling the vent safe.
What to Settle Before the Next Filter Change
Three checks sort most glycol H2S problems. Is the still-column vent going somewhere that can take the actual overhead concentration. Has anyone sampled that overhead at operating temperature, not only the inlet gas. And is the filter-change procedure written for sour glycol. A dehydrator that fails those three will keep putting sulfide into a vent or a fuel line no matter how dry the sales gas is.
Hydrogen sulfide in a glycol system is a side stream that the contactor creates. Treat the overhead, keep the glycol from becoming a sulfide reservoir, and do not use the inlet gas specification as a substitute for a reading at the still column.








