Ice Formation in De-Ethanizers

De-ethanizer columns operate under challenging thermodynamic conditions. Because the overhead product (C2s) requires significantly lower temperatures to condense compared to heavier fractions, the top trays and overhead condensers become highly susceptible to ice and hydrate formation if moisture breaches upstream safeguards.

The Margin Illusion: Ice vs. Hydrates

A common trap in column monitoring is relying solely on the hydrate formation curve. True gas hydrates form when free water condenses in the presence of light hydrocarbons (methane, ethane, ethylene) at specific high-pressure/low-temperature intersections.

Even with a theoretical hydrate margin, parts-per-million (ppm) levels of water escaping the cracked gas molecular sieves can experience localized freezing. When process gas hits localized cold spots—such as condenser tube surfaces or areas experiencing Joule-Thomson cooling across a restriction—water drops below its frost point and transitions directly into solid ice, anchoring to the metal surfaces.

Process Indicators & Symptoms

Early detection relies on recognizing specific dynamic shifts in column behavior rather than waiting for gross failure. Key variables to monitor include:

• Column Pressure Drop (ΔP): A steady, non-linear increase in differential pressure across the top section indicates cross-sectional area restriction in the trays or downcomers.
• Loss of Separation Efficiency: Unexplained increases of C2s in the bottoms (propane/propylene stream) or heavier components slipping into the overhead.
• Condenser Fouling: A drop in the heat transfer coefficient (U-value) across the overhead condensers, requiring increased refrigerant flow to maintain the overhead drum temperature.

Mitigation & Control Strategies

Addressing the formation requires a two-pronged approach: immediate mitigation to maintain operability, and upstream correction to eliminate the moisture source.

• Methanol Injection: Continuous or batch injection of methanol suppresses the freezing point of water and breaks down existing ice/hydrate structures. The injection point must be carefully selected for maximum dispersion.
• Upstream Molecular Sieve Monitoring: Ensure cracked gas dryers are regenerating properly. Monitor the water analyzer trends at the dryer outlet closely; any premature breakthrough is the primary feeder for de-ethanizer icing.

Soft Sensor & Data Integration

To move from reactive to proactive monitoring, OSIsoft PI data can be leveraged. By building a multivariable model (utilizing tools like Aspen ProMV or custom Python scripts), we can calculate real-time theoretical vs. actual heat transfer coefficients and column ΔP. Setting statistical control limits (SPC) on these residuals provides early warning of ice deposition weeks before it impacts plant throughput.