17.4 Inerting, Purging, Discharge Location, and Containment

17.4 Inerting, Purging, Discharge Location, and Containment

The NCEES specification lists Protective systems (e.g., pressure relief, inerting, discharge location, secondary containment) under Safety, Health, and Environment. Section 17.2 covered relief device sizing in depth. This section covers the remaining three, which are examined as short calculations and as judgment questions about where effluent goes and what happens when it gets there.

---

1. Inerting and the Limiting Oxygen Concentration

A flammable mixture needs fuel, oxidant, and an ignition source. Relief and hazard analysis address ignition; inerting attacks the oxidant leg, and it is the most robust of the three because it does not depend on eliminating every possible ignition source.

The controlling parameter is the limiting oxygen concentration (LOC), also called the minimum oxygen concentration: the oxygen level below which no fuel-to-air ratio will propagate a flame.

Two structural facts: the LOC with (\text{CO}_2) is always higher than with (\text{N}_2) because carbon dioxide's larger heat capacity quenches the flame more effectively per mole, so less dilution is needed; and hydrogen's very low LOC is why hydrogen service is so difficult to inert.

NFPA 69 margin requirements. You do not operate at the LOC.

• If oxygen is continuously monitored and the LOC is at least (5%): maintain the oxygen concentration at least 2 volume percentage points below the LOC. For methane (LOC (\approx 12%)), that means operating at or below (10\text{ vol% O}_2).
• If the LOC is below (5%): maintain oxygen at no more than 60% of the LOC. For hydrogen (LOC (\approx 5%)), that means (3\text{ vol% O}_2).

---

2. The Four Purge Methods

Cyclic purging (pressure or vacuum)

Both cyclic methods have the same composition law. Each cycle replaces the vessel contents with inert in proportion to the pressure ratio:

$\boxed{\ y_n = y_0 \left( \frac{P_L}{P_H} \right)^{n}\ }$

• Pressure purging: start at (P_L), pressurize with inert to (P_H) (composition dilutes), then vent back to (P_L) (composition unchanged). Faster, because pressurizing is quicker than evacuating.
• Vacuum purging: start at (P_H), evacuate to (P_L) (composition unchanged), then repressurize with inert to (P_H) (composition dilutes).

The decisive engineering question is mechanical, not chemical. Vacuum purging demands a vessel rated for full vacuum. Atmospheric storage tanks built to API 650 are typically rated for only a few inches of water column of external pressure and will implode if evacuated. Pressure purging demands a vessel rated for (P_H). Choose the method the vessel's design pressure permits.

Worked example. A (10\text{ m}^3) vessel contains air ((21\text{ vol% O}_2)). Pressure purge with nitrogen between (1.0\text{ bar(a)}) and (5.0\text{ bar(a)}). How many cycles reach (1.0\text{ vol% O}_2)?

$\frac{1.0}{21} = (0.200)^n \quad \Longrightarrow \quad n = \frac{\ln(0.04762)}{\ln(0.200)} = \frac{-3.045}{-1.609} = 1.89 \to \mathbf{2\text{ cycles}}$

Verify: after two cycles (y = 21(0.200)^2 = 21(0.0400) = 0.84\text{ vol%}). Cycles must be rounded up; 1.89 cycles is not an operation you can perform.

Sweep-through purging

Inert flows continuously through a well-mixed vessel at constant temperature and pressure:

$y = y_0 \exp\!\left( -\frac{Qt}{V} \right) \qquad \Longrightarrow \qquad V_{\text{inert}} = Qt = V \ln\!\left( \frac{y_0}{y} \right)$

For the same vessel going from (21%) to (1%): (V_{\text{inert}} = 10\ln(21) = 30.4\text{ m}^3) of nitrogen at operating conditions. At (Q = 0.50\text{ m}^3\text{/min}) that takes (60.9\text{ minutes}). Sweep-through is the only method available for a vessel that can tolerate neither vacuum nor pressure — which describes most atmospheric tanks — but it consumes inert continuously and requires a genuine flow path in and out, not a dead-ended nozzle.

Siphon purging

Fill the vessel completely with liquid (often water), then drain it while feeding inert into the vapor space. The inert requirement equals the vessel volume at essentially atmospheric pressure, making it the cheapest method by far for large vessels. The constraint is that the vessel must be able to carry the hydrostatic load of a full liquid fill and must tolerate the liquid.

Blanketing (padding)

Distinct from purging: a continuous low-pressure nitrogen supply holds a slight positive pressure on a storage tank vapor space, admitting nitrogen on inbreathing (pump-out or cooling) and venting on outbreathing (fill or solar heating). API 2000 governs the venting capacity calculation. Blanketing prevents air ingress, suppresses oxidative product degradation, and keeps the vapor space permanently below the LOC.

---

3. Discharge Location: Where Relief Effluent Goes

Sizing a relief valve correctly and then discharging it into an unsafe location is a complete design failure. API 521 governs the decision.

When atmospheric discharge is permitted, the tailpipe rules exist because each has killed someone:

• Vertical and upward, never horizontal and never angled toward a platform, walkway, or air intake. The discharge must be high enough above working areas that the jet disperses before it reaches anyone.
• No rain cap. A rain cap can freeze, corrode shut, or be blown off as a projectile, and it deflects the jet sideways.
• Weep hole at the low point of the tailpipe. Without it, rainwater accumulates and is expelled as a liquid slug on the next relief event, imposing a large reaction force on the valve and piping.
• Reaction force must be carried by the tailpipe supports, not by the valve body or the vessel nozzle.

Flare systems add their own requirements: a knockout drum to remove liquid before the flare tip (burning liquid rains fire), a liquid seal drum and purge gas to prevent air ingress and flashback into the header, and a continuous pilot with flame detection.

---

4. Secondary Containment

Secondary containment catches what primary containment releases. For bulk liquid storage, the EPA SPCC rule (40 CFR 112) requires containment capacity for the entire capacity of the largest single container plus sufficient freeboard for precipitation; the common design practice is to size the dike at 110% of the largest tank, not the sum of all tanks in the dike.

Design points the exam can test:

• Sizing basis: largest single tank, not total inventory. Simultaneous failure of two tanks is not a credible design case, but simultaneous failure plus a rainstorm is.
• Volume displacement: the volume of the other tanks standing inside the dike below the design liquid level reduces the dike's usable capacity and must be subtracted.
• Drainage control: the dike drain valve must be normally closed and under administrative control, so that accumulated rainwater is inspected before release rather than draining automatically.
• Impermeability and compatibility: earthen dikes fail for strong solvents; concrete or lined containment is required where the liquid attacks the substrate.
• Spacing and slope: floors slope away from the tank shell toward a remote impounding area so that a burning spill does not sit against the tank it came from.