The tower pressure is the originating signal. The protection path then runs through three pressure switches, trip logic, a solenoid, a 6-inch HV, the high-pressure nitrogen line, the tower vapor space, and finally the pressure relief valve. Follow that path in order. The present conflict is physical: nitrogen supplied above 4 barg can continue entering a tower rated at 3 barg after the vacuum has been removed, causing the relief valve to lift.
Where does the protection path stop controlling pressure?
| Path element | Reading or observation | Decision |
|---|---|---|
| Tower | Pressure-versus-time trend during a steam trip | Establish the vacuum formation rate before nitrogen admission. |
| Three pressure switches | Individual states and calibrated trip/reset pressures | Determine whether all channels detect the event and whether the voting logic acts as intended. |
| Logic and solenoid | Command and output timestamps | Separate sensing delay from logic, solenoid, and valve delay. |
6-inch HV |
Stem position or independent open/closed indication | Confirm that the commanded position matches the physical position. |
| Nitrogen supply | Pressure upstream of the valve and tower pressure | Quantify the driving pressure across any valve, regulator, or restriction. |
| Relief valve | Lift indication and tower pressure at lift | Identify whether excessive flow, late closure, or both caused the demand. |
A vacuum trip only initiates admission. It does not, by itself, limit the final positive pressure. If the HV stays open until a separate reset condition occurs, the large pressure differential can drive nitrogen into the tower after pressure recovery. The relief valve then becomes the stopping device, which explains repeated lifting during events such as a steam trip.
Is the physical system capable of surviving the vacuum transient?
Layer one first: establish the tower's mechanical pressure envelope before changing control hardware. The recorded rating is 3 barg, while no vacuum rating is recorded. A mechanical rerating calculation may find allowable vacuum capacity, but operating practice must use the approved result rather than an assumed maintenance rating.
| Mechanical result | Meaning | Next check |
|---|---|---|
| No allowable vacuum | The protection must arrest pressure decline before the tower crosses atmospheric pressure by the permitted amount. | Measure the fastest credible pressure-decay rate and total response time. |
| Limited vacuum rating approved | The rating supplies a quantified response margin. | Compare the minimum transient pressure with that approved limit. |
| Full-vacuum capability approved | Loss of nitrogen admission may no longer be the governing collapse case. | Reassess whether the existing trip remains necessary and what hazards nitrogen admission creates. |
Use calibrated absolute pressure for collapse analysis. Gauge pressure alone can obscure the distance to vacuum, particularly when comparing archived events taken under different atmospheric conditions.
Do the switch and valve settings create the PSV lift?
Trend each switch, the voted trip output, solenoid output, HV position, tower pressure, and relief indication on one time base. Do not infer valve motion only from the solenoid command. A sticking valve, incorrect fail action, blocked sensing connection, or wide switch reset band can produce the same pressure trace as excessive valve capacity.
| Observed result | Likely stopping point | Action |
|---|---|---|
| Switches change late | Pressure measurement or switch calibration | Inspect sensing connections and calibrate each switch. |
| Trip output changes promptly but HV opens late | Solenoid, actuator, or valve mechanics | Measure command-to-position travel time and repair the slow element. |
| HV opens promptly but never receives a close command | Reset logic or missing positive-pressure cutoff | Add a defined closure function or a self-limiting pressure device. |
| Close command occurs but pressure keeps rising | Valve leakage, slow closure, or stored downstream nitrogen | Test shutoff leakage and trend position through the full stroke. |
| HV closes correctly but the relief valve still lifts | Admission rate exceeds the useful pressure margin | Size a restriction or staged regulating path from transient data. |
A candidate regulator setting of 1 barg or 2 barg was proposed, but neither value is a design setpoint without the relief-valve set pressure, accumulation basis, regulator lock-up behavior, sensing accuracy, and transient margin. The vessel rating of 3 barg is not automatically the relief set pressure.
Which protection branch resolves both hazards?
| Branch | Vacuum response | Positive-pressure behavior | Reliability decision |
|---|---|---|---|
| Existing switches and full-flow HV | Fast when the complete chain operates | Can lift the relief valve if flow continues | Verify voting, reset logic, response time, and proof-test coverage. |
| Fixed restriction in nitrogen line | Slower; capacity must exceed the worst vacuum-generation rate | Limits pressure-rise rate but does not provide a pressure stop | Check blockage, erosion, upstream pressure range, and credible bypass paths. |
| Self-contained regulator | Admits nitrogen while downstream pressure is low | Closes near its controlled pressure, subject to droop and lock-up | Manufacturer qualification, failure position, sensing arrangement, and proof testing determine whether it can carry a safety function. |
| Pilot-operated device sensing tower pressure | Can provide self-actuated admission | Can stop flow at a selected tower pressure | Confirm that the device is designed and documented for this inlet-service configuration. |
| Parallel operating-control path | Acts first during routine events | Stops normal admission before relief lift | Keep the independent protection path available for failure of the operating path. |
| Mechanical rerating | Increases tolerable vacuum if approved | Does not limit nitrogen overpressure | Use the approved pressure-temperature envelope in the hazard review. |
Ambient-air vacuum breaking is unsuitable when the process requires inerting because of flammability and temperature. Nitrogen remains the applicable admission medium. A standard process regulator is not automatically equivalent to a relief valve in probability of failure on demand. Assigning it a safety role requires a defined target, documented failure modes, independence from initiating causes, proof-test intervals, diagnostic coverage, and common-cause assessment.
How should the restriction or regulator be sized?
Use an archived event before the nitrogen valve opened. From tower volume and the absolute pressure-versus-time trace, calculate the gas inflow needed to arrest the decline. For a fixed volume with approximately constant gas temperature, the molar-rate estimate is:
dn/dt = V/(R × T) × dP_abs/dt
If temperature changes materially, include the temperature term from n = P_abs × V/(R × T). Do not fit the restriction from gauge pressure or from the pressure-rise portion after nitrogen admission.
- Extract the fastest credible pre-admission pressure decline and the corresponding temperature.
- Calculate the minimum nitrogen rate needed to keep pressure above the approved vacuum limit.
- Calculate restriction or regulator capacity across the full upstream supply range, including the supply stated to exceed 4 barg.
- Model the recovered-pressure portion with the proposed flow limit and valve closing behavior.
- Confirm that the predicted minimum pressure stays inside the vacuum limit and the predicted maximum remains below relief-valve operation with engineering margin.
- Check failure cases: regulator wide open, restriction bypassed, sensing line blocked, HV failing open, and loss of nitrogen.
What commissioning procedure proves the resolving branch?
- Approve the tower's allowable vacuum and positive-pressure limits through the mechanical review.
- Calibrate all three switches and record their trip and reset readings individually.
- Verify the actual voting logic, solenoid fail state, HV stroke direction, travel time, and shutoff leakage.
- Install the selected restriction, regulator, pilot device, or staged operating path using the reviewed capacity calculation.
- Test each protection channel and each credible single failure without defeating the independent relief function.
- Run a controlled transient, recording absolute tower pressure, temperature, switch states, logic output, solenoid command, HV position, nitrogen pressure, and relief indication on one synchronized trend.
FAQ
Can I use a pressure regulator as the tower vacuum safety device?
Only after its failure action, capacity, sensing arrangement, qualification, diagnostics, and proof testing meet the assigned probability-of-failure target. A self-contained process regulator does not inherit relief-valve reliability merely because it controls pressure.
Can I install an orifice to stop the relief valve lifting?
An orifice can limit nitrogen flow, but size it from the tower volume and pre-admission absolute pressure decay. Verify both branches: enough flow to avoid the approved vacuum limit and little enough flow to avoid relief operation.
Does a 3 barg tower rating define the regulator setpoint?
No. Select the setpoint from the approved vessel envelope, relief set pressure, regulator droop and lock-up, instrument error, and transient margin; proposed values of 1 barg or 2 barg are only candidates.
Can I use ambient air instead of nitrogen?
Not where the service requires inerting because of flammability and temperature. Use the specified inert medium and include loss of its supply in the protection review.
Does switch operation prove the vacuum protection works?
No. The final verification is a synchronized controlled-transient trend showing that the switches, logic, solenoid, and HV respond in sequence while tower pressure remains above the approved vacuum limit and below relief-valve operation.