Sound velocity vs conductivity and photoelectric interface detectors
Three ways to make the same call. The difference is not accuracy but the premise each one rests on. This page sets the three side by side: discriminator, premise, behaviour on a clear pair such as dichloromethane and water, response to fouling and changeover, and which way each fails.
Feature comparison
| Feature | Inline Ultrasonic Phase Separation Detector | 行业通用 Industry-standard 电导式 / 光电式分层仪 |
|---|---|---|
| Discriminator | Sound velocity (time of flight) + echo amplitude | Conductivity switching / transmittance or refraction |
| Premise | Phases differ by ≥ 30 m/s; the line runs full | Conductivity: aqueous conducts, stable, electrode may wet. Photoelectric: an optical difference, a clean window |
| Dichloromethane / water | 390 m/s apart — decided directly | Conductivity works; turbidity/colour units fail (both phases clear), refractive units work |
| Aqueous conductivity moves | Not part of the discriminator | Conductivity threshold drifts and needs resetting |
| Dark or turbid liquor | Not part of the discriminator | Transmittance units fail; refractive units suffer from fouling |
| Rag layer | A third state: 6 dB below baseline reports emulsion and holds the valve | Most units are two-state; the rag layer falls to whichever side the threshold sits |
| Wetted parts and tapping | Inline spool, flush sapphire windows | Conductivity needs an inserted electrode; photoelectric needs a window |
| PTFE-lined line | Lined spool, windows through the liner | Electrode tapping is awkward; a sight glass needs its own opening |
| Fouling | Amplitude baseline drops and alarms first; sapphire cleans with a rinse | Deposit on an electrode or window changes the decision, often silently |
| Product changeover | Re-sample both phases and reset two bands, about five minutes; eight stored recipes | Conductivity thresholds usually need resetting; optical difference must be re-confirmed |
| Failure direction | Toward not-decidable → hold the valve and alarm | Most units give a confident wrong answer, silently |
| Where the cost sits | Above a conductivity switch, below inline spectroscopy or refractometry | Conductivity lowest; photoelectric in between |
Takeaway
A conductivity switch is the cheapest answer wherever the aqueous phase conducts and an electrode may be fitted, which covers most interface duties — no reason to replace it for the sake of a newer principle. A refractive sight-glass unit is equally direct where the indices are far apart and the window stays clean.
Sound velocity owns three situations: both phases clear and colourless, so colour and turbidity methods fail; both non-conducting, or an aqueous conductivity that moves, so the conductivity premise softens; or a duty that needs the rag layer treated as its own state with a conservative interlock. Dichloromethane and water hit the first two at once, which makes that pair close to a textbook case for the method.
The cost is on the table: a two-phase calibration on site, repeated at changeover; a line that must run full; and no answer at all for pairs closer than 30 m/s. Where those three do not hold, the other two methods fit better.