In nuclear glove box work, the negative pressure gradient is not a comfort setting. It is the containment barrier. Nuclear glove box negative pressure gradient design determines how air moves across every boundary, so that a breach draws air inward instead of pushing contamination out. The three-stage cascade below covers the room-to-chamber, chamber-to-lock, and lock-to-exhaust boundaries, together with the leak rate acceptance tests that prove the design actually holds.
Nuclear Glove Box Negative Pressure Gradient Design: The Three Boundaries
The cascade must be monotonic. Each stage closer to the exhaust and further from the operator is held at a progressively lower gauge pressure. That ordering is what keeps airflow direction predictable no matter which seal fails first.
Stage one is the room-to-chamber boundary. The operating room typically sits between -10 and -30 Pa relative to the adjacent corridor, and the glove box chamber between -80 and -150 Pa. This 60 to 120 Pa step is usually the largest in the facility, because gloves and main door seals are the dominant leak paths.
Stage two is the chamber-to-transfer-lock boundary. The transfer lock, waste port, or pass-through is held 50 to 100 Pa below the main chamber, commonly in the -180 to -250 Pa range. When the outer door opens to the room, air flows in rather than out. Reversing this step, a frequent error in retrofits, turns the lock into a buffer that discharges toward the operator on door opening.
Stage three runs from the lock or chamber to the exhaust plenum. That plenum and the HEPA train sit 100 to 200 Pa below the most negative enclosure, which keeps filter housings, duct joints, and dampers under suction at all times.
| Boundary | Typical step | Purpose |
|---|---|---|
| Room to chamber | 60-120 Pa | Contains glove and seal leakage |
| Chamber to transfer lock | 50-100 Pa | Contains door-opening transients |
| Lock to exhaust plenum | 100-200 Pa | Keeps the HEPA train under suction |
Setting and Verifying the Three Pressure Steps
Do not copy these numbers directly from any table. The safety analysis report and the facility ventilation design govern the values, and the figures above are only a sanity check during review.
Instrumentation matters as much as the setpoints. Reference differential pressure transmitters to the corridor or a dedicated reference node rather than the room static pressure, protect the impulse lines from dust and condensation, and calibrate to within 1 Pa. A 5 Pa drift on a 50 Pa step is a 10 percent error and will fail a gradient verification even though the control loop looks healthy.
Control response should be fast enough to hold the cascade when a glove loosens or a door is cracked. Damper or VFD control that settles in 2 to 3 seconds is adequate for most boxes; slower loops let chamber pressure drift toward room pressure during transients. Fail-safe logic should close the supply damper and drive the exhaust fan to full speed on loss of the exhaust interlock, with the high-pressure alarm at 70 percent of setpoint so operators have time to react.
Leak Rate Acceptance and Test Sequence
A gradient that reads correctly on a gauge proves nothing about the leak rate of the enclosure itself. Every nuclear glove box negative pressure gradient design should therefore be accepted through a global pressure-decay test, followed by local helium testing of welds and seals, and finished with a live verification of the three-stage cascade with gloves and doors installed.
For the global test, isolate the enclosure, pull it to the design negative differential, commonly -1000 Pa or the test pressure stated in the design basis, and record the decay over 30 to 60 minutes. Report the result as a fractional volume loss per hour and correct it for barometric and temperature change, or the number is meaningless. ISO 10648-2 defines the tightness classes and the pressure-change method; most nuclear installations specify Class 2 or better, which in practice lands near 0.1 to 0.5 vol%/h at the test differential. Helium mass-spectrometer testing of window seals, glove rings, and weld joints typically targets 1 × 10⁻⁶ atm·cc/s or tighter.
Three acceptance details are routinely missed. Test with the gloves installed and the ports in their operating configuration, because gloves are usually the largest single contributor to measured leak rate. Run the decay test at the specified test differential rather than a lower pressure that is easier to hold. Verify the running gradient at design exhaust flow, since a static test says nothing about pressure stability under dynamic leakage.
If any boundary in a nuclear glove box negative pressure gradient design is non-monotonic, the cascade collapses into a single barrier and the containment argument fails. Two practical rules cover almost every failure mode. Confirm the pressure order from room to exhaust before tuning individual setpoints, and treat the pressure-decay result and the running gradient as separate acceptance tests that both must pass.
