Hydrate Control in Subsea Gas Flowlines: Methanol versus MEG Injection

The operating problem

A subsea gas flowline that drops into the hydrate stability region can plug. Hydrates need free water, gas, high pressure and low temperature all at once. On a long tieback the fluid cools toward seabed temperature, and every shutdown and restart moves the system closer to the hydrate envelope. A plug costs production, makes remediation hard and, if mishandled, can become a safety event. On subsea gas, insulation and heating seldom keep a long offset line outside the envelope on their own, so most operators inject thermodynamic hydrate inhibitors (THIs). Methanol (MeOH) and monoethylene glycol (MEG) are the two main choices.

The choice shapes the whole development. It drives umbilical or injection-line sizing, topsides storage, the regeneration plant, export-quality risk and operating cost for the life of the field. This article compares the two on those terms.

Standards and requirements context

Nothing in the standards picks MeOH or MEG for you. They frame the decision, and the choice has to be justified inside that frame:

• ISO 13623 covers pipeline transportation systems for the petroleum and natural gas industries. It sets the design and operating basis, including how internal fluid hazards get managed.
• DNV-ST-F101 covers submarine pipeline systems. It sets the integrity and operability context for the flowline.
• ISO 13628-1 gives general requirements and recommendations for designing and operating subsea production systems, including chemical injection.
• ISO 13628-5 covers subsea umbilicals. It tells you which injection-line materials, diameters and pressure ratings you can actually get.

The hydrate management philosophy (prevention, margin to the hydrate curve, shutdown and restart strategy) lives in your project flow-assurance basis of design. Treat the standards as the integrity envelope, not as inhibitor selection rules.

How both inhibitors work

Both inhibitors work the same way. They dissolve into the free water phase and push the hydrate equilibrium curve to lower temperature (or higher pressure), which keeps operating conditions outside the hydrate stability zone. The mechanism is shared. The operating profiles are where they differ.

For a first look at how much suppression you'll get, people often reach for the Hammerschmidt relation:

ΔT = K·W / (M·(100 − W))

Here ΔT is the hydrate temperature suppression, W is the inhibitor concentration in the aqueous phase (wt%), M is the inhibitor molecular weight and K is an empirical constant for the inhibitor. Use it for screening only. Final design needs a validated hydrate prediction package run on the actual gas composition, water chemistry and pressure range. The simple relation loses accuracy at high inhibitor concentration and in systems with salts.

Comparing methanol and MEG

Where each inhibitor goes

Methanol partitions strongly into the gas and, to a lesser degree, the condensate. So some of what you inject never reaches the water phase where it's needed. It leaves with the export gas and liquids and can't be economically recovered. That loss sets the operating cost, and depending on processing and sales terms it can also cause specification or penalty trouble downstream.

MEG barely vaporises, so nearly all of it stays with the produced water. What comes back is "rich" MEG, diluted by produced and condensed water, and it can be regenerated and reinjected. The GATE Energy summary notes that continuous MEG injection is common on long offset gas flowlines, mainly because MEG is easier to recover and reuse than methanol.

Side-by-side comparison

The 2005 OTC paper (OTC 17355) compares MEG with methanol using experience from four major Norwegian gas/condensate developments. It treats the MEG injection and recovery technology as a central part of the decision. The lesson is the usual one: the more continuous the inhibitor demand and the longer the line, the more the regeneration plant pays for itself.

Injection rate

Water rate sets the injection rate. Hydrocarbon rate matters much less. For a first mass balance, look at the aqueous phase leaving the injection point. If the lean inhibitor is neat (mass fraction of 1) and you ignore losses to other phases:

q_i = q_w · x / (1 − x)

where q_i is the inhibitor mass rate, q_w is the water mass rate (produced plus condensed) and x is the required inhibitor mass fraction in the aqueous phase. For lean MEG that already carries water, swap the "1" for the lean mass fraction x_lean:

q_i = q_w · x / (x_lean − x)

Three things follow:

• The required x comes from the hydrate curve at worst-case pressure and temperature, plus whatever margin your flow-assurance philosophy calls for.
• The rate climbs steeply as x approaches x_lean, so every point of margin on x costs real injection volume.
• For methanol, add the amount lost to gas and condensate. That's an equilibrium calculation in the process simulator, and at low water rates it's often the biggest term.

Water rate moves over field life. Early on you may only have condensed water; later, formation water shows up. Size pumps and lines for the late-life case, and make sure turndown holds up for the early case. Also confirm the injection quill actually delivers inhibitor so it mixes with the water along the full flowline. The Norwegian University thesis "Rate of Hydrate Inhibitor in Long Subsea Pipelines" tackles this distribution question for long lines, and it matters on any long tieback.

Regeneration requirements

MEG regeneration is where the capital goes. A rich MEG stream typically needs:

• Water removal, to bring the glycol back to the lean concentration the injection rate was calculated from.
• Salt management. Formation water brings dissolved salts, and they concentrate as water boils off. You need a reclamation step to prevent scaling and fouling, and this is the part of the plant most often undersized.
• Removal of hydrocarbons and solids carried in the rich stream before the regeneration column.
• Buffer storage for lean and rich MEG. You need that inventory to restart after an upset.

Methanol needs none of this when used once-through. So the comparison is really loss cost and export impact on one side, against the capex and operability burden of the regeneration plant on the other.

Practical example: methanol startup dosing

The Scribd document on the KG 98/2 gas field project describes methanol injection at 2 m³/h for well startups, and for operation when temperatures drop below 4 °C. Methanol is stored on the central processing platform and injected from there. This is intermittent, event-driven use, and a regeneration plant wouldn't be justified for it.

The consumption arithmetic is simple:

2 m³/h × 24 h/day = 48 m³/day

Multiply any campaign length through that figure to size storage. For a typical restart, the real question is how long the system sits in the hydrate-prone condition. That's a flow-assurance result, not a number you can pull from a general rule.

Illustrative scenario (not from a listed source): take two gas fields. One has a short tieback and occasional cold restarts. The other has a long offset line, steady water production and a water cut that climbs as the field matures. On the first, methanol dosing at startup and shutdown keeps topsides simple. On the second, once-through methanol means a chemical bill that keeps growing and a continuous export-quality issue, so a MEG loop is the more defensible call despite the regeneration plant.

A recent JPT article, "Flow-Assurance Strategy Manages Subsea Asset Without Continuous MEG Injection," shows continuous injection isn't the only route. Operating, monitoring and restart philosophies can cut inhibitor demand. Work through those options before sizing a regeneration plant for the maximum case. The JPT piece on MEG in the Gulf of Mexico treats MEG as an established deepwater gas option too, which supports considering it wherever recovery infrastructure exists.

Decision checklist

• Define the hydrate-prone operating envelope across steady state, shutdown, cooldown and restart, using a validated hydrate prediction.
• Establish the water rate profile over field life (condensed water first, then formation water) and size for both ends.
• Calculate the required inhibitor concentration in the aqueous phase, add margin, then calculate injection rate from the mass balance, including partition losses for methanol.
• Decide whether demand is continuous or intermittent. Intermittent points to methanol; continuous points to MEG.
• Check line length and the cold-temperature behaviour of the injection fluid, especially MEG viscosity in the umbilical.
• For MEG, confirm the regeneration and reclamation design basis for salts, hydrocarbons and lean concentration control.
• For methanol, confirm export gas and condensate specifications, downstream processing limits, and handling and storage hazards.
• Check injection point design, quill mixing and the turndown control strategy.
• Review whether operating strategies reduce inhibitor demand before sizing for the maximum case.
• Cover safety for any equipment that must be opened: isolate, depressurise, verify zero energy, apply lock-out/tag-out, observe hazardous-area precautions, use gas detection and vent to a safe location. For a suspected hydrate plug, treat both sides as potentially pressurised. Use a specialist remediation procedure, and never depressurise one side only.

Conclusion

Methanol and MEG suppress hydrates by the same mechanism, so the decision comes down to how often you need the inhibitor, how long the line is, and whether you can get the inhibitor back. Methanol suits short, intermittent duty where simplicity wins. MEG suits long, continuous duty, where the recovery plant offsets chemical losses and export impact.

Next steps: build the water-rate profile over field life, run the hydrate prediction for the worst case, and calculate injection rates for both inhibitors from the mass balance, including methanol partition losses. Then compare the lifetime cost of methanol losses against the capex and operability burden of MEG regeneration, and test whether an operating strategy can cut demand before you commit to either.