Skip to content

Phase-Boundary Monitor

Monitor / dashboard — instantiates Controlled Demixing and Domain Formation

Tracks how close the running system sits to its phase boundary and raises an alarm the moment it crosses into — or out of — a separated state.

A Phase-Boundary Monitor is the live instrument that answers, continuously, how far are we from separating right now? It overlays the system's current state — temperature, composition, pressure, whatever the relevant axes are — onto the known coexistence and instability boundaries, and reports the margin to each. Its defining focus is thermodynamic proximity, not morphology: it is not measuring how big the domains are, it is watching whether the system is about to enter (or has unexpectedly entered or left) a region where domains form at all. When the margin narrows or a boundary is crossed, it signals — turning a silent, invisible thermodynamic drift into an event someone can respond to before the batch clouds.

Example

An operator running a waxy crude through a chilled pipeline segment watches a phase-boundary monitor keyed to the wax appearance temperature — the boundary below which paraffins begin to precipitate and coat the pipe wall. The dashboard reads the live fluid temperature and composition against that boundary and displays the margin: ≈6 °C above onset, narrowing. As ambient cooling deepens overnight, the margin shrinks; at a set threshold the monitor alarms, well before any actual deposition, giving the operator time to raise the inlet temperature or inject a flow improver.

Later, an unexpected composition change from a new feed shifts the boundary itself upward, and the live state — unchanged in temperature — suddenly sits below onset. The monitor catches the crossing not because temperature moved but because the boundary did, and flags an unexpected entry into the precipitating region. In both cases the value is the same: the separation was made visible as a tracked distance, and acted on as a margin, rather than discovered as a blockage.

How it works

The monitor consumes a boundary map (where coexistence and instability lines sit in the relevant state space) and a live read of the system's current coordinates, and computes the signed distance between them — the separation criterion, expressed as a margin. What distinguishes it from a general process dashboard is that its reference is the phase boundary itself: alarms fire on proximity to and crossings of that line, in either direction, including cases where the boundary moves under a composition shift while the operating point holds still. It runs as a closed loop — sample, locate, compare, signal — for the duration of the process.

Tuning parameters

  • Alarm margin — how much cushion to the boundary triggers a warning. A wide margin gives early notice but cries wolf; a tight one is quiet until you are nearly across.
  • Sampling cadence — how often the state is read. Fast sampling catches sudden excursions but adds noise and cost; slow sampling can miss a quick dip across the line.
  • Boundary source — a fixed reference line versus one that is itself updated live as composition changes. A moving boundary catches feed-driven crossings but depends on trustworthy composition data.
  • Directionality — whether the monitor watches only for entering the separated region, only for leaving it (unwanted re-dissolution), or both.
  • Hysteresis band — how far the state must recover before an alarm clears, to avoid chattering when the margin hovers on the line.

When it helps, and when it misleads

Its strength is converting an invisible thermodynamic approach into a leading indicator: it warns before separation is visible, catches boundary crossings caused by the boundary moving rather than the state, and works in both directions so it flags both unwanted demixing and unwanted remixing. It is the run-time complement to the design-time maps.

It misleads chiefly by trusting a stale or wrong boundary: the margin is only as good as the coexistence map it references, and a boundary measured under clean, equilibrium conditions can sit in the wrong place for a real, kinetically sluggish, additive-laden process — so the monitor can read "safe" while deposition has already begun in a metastable pocket.[1] It also reports proximity, not consequence: being just past the boundary may be harmless or ruinous depending on kinetics it cannot see. The discipline is to validate the referenced boundary against the actual system, carry a margin sized to that uncertainty, and pair the monitor with a morphology measurement that confirms whether separation is truly underway.

How it implements the components

  • stability_margin_and_separation_criterion — its core output: the live signed distance from the current state to the coexistence or instability boundary, and the rule for when that counts as "separated."
  • phase_morphology_monitoring_loop — it operationalizes the ongoing sense-and-signal loop, sampling state and firing on entry, exit, and near-approach for the whole run.

It does not produce the boundary map it watches against — that is Phase-Diagram Mapping — and it does not measure the domains once they form; sizing and connectivity are Domain-Morphology Imaging. The monitor tracks proximity, not structure.

References

[1] Between the binodal (equilibrium coexistence) and the spinodal (limit of metastability) lies a region where a mixture is thermodynamically able to separate but kinetically may not, until a nucleation event tips it. A boundary monitor keyed only to the equilibrium line can therefore read "clear" inside this metastable band while separation quietly begins — which is why the referenced boundary must be validated against the real system. ↩