Stable control comes from making the level measurement reject headspace-pressure changes and represent liquid inventory rather than the moving foam surface. For this boiling hydrochloric acid tank at 236 °F, correct the bubbler dynamics first; if that cannot produce a stable, repeatable signal, move to true differential-pressure measurement with chemically compatible remote seals. Look at the trend first. Tuning does not fix wiring, unequal impulse paths, density variation, or an undefined measurement target.
What do the symptoms say about the signal chain?
Separate actual inventory movement from measurement error by trending the level signal, headspace pressure, feed-valve command, and acid feed together. A level indication that moves immediately with headspace pressure while feed and inventory remain nearly unchanged points to incomplete pressure compensation or unequal pneumatic response. A slower level change that follows heat input, boiling intensity, or composition may come from changing liquid density or vapor void fraction.
| Signal | Source | Wrong-value symptom |
|---|---|---|
| Measured differential pressure | Hydrostatic head plus purge and headspace-pressure effects | Fast spikes or zero shifts during vessel-pressure changes |
| Headspace pressure | Boiling, vapor generation, and pressure control | Correlation with indicated level reveals failed dynamic cancellation |
| Feed-valve command | Level controller output | Rapid hunting shows the controller is reacting to measurement noise |
| Boiling intensity | Heat input and process conditions | Changes can alter foam, density, and liquid-vapor void fraction |
| Independent inventory indication | Mass balance or a suitable reference measurement | Disagreement identifies whether the DP signal represents actual inventory |
Decide what the loop must control: continuous liquid inventory or the top of the froth and splash zone. DP measures hydrostatic liquid head and cannot reliably identify the top of foam. Ultrasonic and radar instruments have already failed in this installation because vapor, foam, and surface agitation disrupt the return signal, so replacing one surface sensor with another does not address the demonstrated failure mode.
Why does the bubbler indicate false level changes?
A correctly referenced DP system calculates level from ΔP = ρgh. Headspace pressure should act on both sides and cancel. In the present bubbler arrangement, one pneumatic path terminates in vapor and the other in liquid. When vessel pressure changes, the gas in both paths compresses, but different tube lengths, internal volumes, flow resistances, purge rates, and endpoint conditions make the paths respond at different speeds. The vapor-side path can respond first, creating a temporary differential pressure that looks like a level change.
Adjusting purge flows may zero the instrument at one steady vessel pressure without correcting that dynamic mismatch. The zero then shifts when pressure changes. Nitrogen injection that stabilizes headspace pressure can reduce the disturbance, but it does not repair unequal sensing paths and adds another control interaction.
Boiling introduces a second mechanism. Hydrostatic pressure depends on density, while vapor bubbles in the liquid create a variable void fraction. Even perfect headspace-pressure cancellation cannot convert DP to geometric level accurately when effective density changes. Filtering can suppress fast pressure transients, but it cannot correct a sustained density error.
Which measurement architecture fits this service?
Retain the bubbler when its wetted tube is suitable for hydrochloric acid, purge paths can be balanced, and the required variable is average liquid inventory. Use signal damping only after correcting the installation and confirming that filtered response remains fast enough for safe inventory control.
If pressure cancellation remains poor, use a true DP transmitter connected between the lower liquid point and vessel headspace. External diaphragm seals isolate the sensing element from the corrosive process and plugged impulse lines. A Rosemount 3051 DP transmitter with external diaphragm seals and at least 2-inch flanged connections was identified as a candidate architecture, but the product family is not a complete specification. Select diaphragm, flange, gasket, fill-fluid, pressure rating, and temperature capability from the actual acid concentration, contaminants, operating pressure, and 236 °F process temperature. Welded tantalum diaphragm options exist, but material compatibility must be checked for this exact hydrochloric acid service.
A stilling well or parallel chamber can calm surface motion, but it must exchange liquid and vent vapor without restriction. A chamber containing cooler acid may reproduce elevation while producing a different density; its DP-derived level can then disagree with the boiling vessel. It also does not inherently cancel changing headspace pressure.
How do I correct the measurement and control loop?
- Define the controlled boundary. State whether the required measurement is liquid inventory or foam height. Use DP for inventory; select a different measurement principle if foam height is the process limit.
- Capture synchronized trends. Record level PV, headspace pressure, feed-valve command, acid feed, and a measure of boiling intensity. Do this before changing controller tuning or damping.
- Inspect both pneumatic paths. Compare tubing length, bore, volume, routing, restrictions, condensate accumulation, purge hardware, leaks, and blockage. Remove avoidable differences between high- and low-side response.
- Check purge behavior. Verify stable purge flow and confirm that the dip tube remains clear. Recheck zero at more than one normal headspace pressure; a zero that moves with pressure identifies unequal dynamic response.
- Apply limited damping. Increase transmitter or control-system filtering only enough to remove pressure-induced fluctuations. Do not filter so heavily that a real inventory excursion is hidden.
- Evaluate true DP if needed. Compare a bottom-to-headspace remote-seal installation against the corrected bubbler. Review chemical compatibility, thermal effects, capillary routing, maintainability, and the required measurement span with the manufacturer.
- Retune last. Place the loop in a controlled operating state, confirm a stable PV, and then tune the level controller. A valve that hunts on a noisy PV needs measurement correction before gain or reset changes.
How do I verify that the correction worked?
Challenge the measurement with normal headspace-pressure movement while holding actual inventory as steady as the process permits. A corrected system should show little immediate level response to pressure alone. Then introduce or observe a known inventory change and confirm that the PV moves in the correct direction without excessive delay.
Compare indicated level against an independent mass balance using measured feed and withdrawal. Repeat the comparison at different boiling intensities. If the error changes with boiling but not headspace pressure, investigate density and void fraction rather than pneumatic compensation. Finally, run the loop in automatic and verify that valve motion follows real inventory demand instead of rapid PV fluctuations.
Which pitfalls recur in boiling acid level service?
Do not treat foam height as liquid level, or calibrate a calm, cool chamber as though its density matches boiling acid. Do not zero unequal purge paths at one pressure and call the problem solved. Avoid using controller tuning to suppress transmitter noise; aggressive tuning transfers that noise directly to the feed valve.
Remote seals also require engineering. Different capillary temperatures, unsuitable fill fluid, poor mounting, and incompatible wetted materials can create new errors or failures. A larger flange may reduce plugging risk, but connection size alone does not establish chemical or thermal suitability.
Frequently asked questions
How do I tell whether vessel pressure is causing the false level?
Trend headspace pressure and level PV on the same time base. Immediate matching movements, without corresponding feed or inventory change, identify incomplete pressure cancellation.
How do I stop a bubbler level signal from hunting?
Match the two pneumatic paths, stabilize purge flow, clear restrictions and condensate, and verify zero at multiple operating pressures. Add only enough filtering to remove the remaining fast fluctuation.
How do I measure level through hydrochloric acid foam?
First decide whether the target is liquid inventory or foam top. DP can indicate hydrostatic liquid inventory but does not directly detect the top of a variable froth layer.
How do I use a stilling well on a boiling acid tank?
Provide unrestricted liquid communication and vapor venting, then verify that chamber temperature and density track the vessel. A cooler chamber can show the right elevation but the wrong DP-derived level.
How do I select diaphragm seals for this service?
Give the manufacturer the acid concentration, contaminants, 236 °F temperature, operating pressure, connection requirements, and required span. Obtain written confirmation for diaphragm, gasket, fill-fluid, flange, and capillary suitability.
Stop field adjustment if the transmitter cannot hold zero across normal pressure changes, the wetted-material compatibility is unresolved, or the required pressure and temperature ratings are unavailable. Escalate to the transmitter manufacturer's official engineering support with process chemistry, pressure, temperature, connection, span, and trend data so it can approve the complete measurement assembly.