储罐区工程

Frank, T-304 — our 80,000 m³ external floating roof crude tank — is due for its API 653 out-of-service inspection this year. Before we drain it, the operators are reporting a strong hydrocarbon smell around the rim seal. That means we're losing vapors. Can you do a preliminary inspection while the tank is still in service?

I'll walk the roof and inspect the primary and secondary rim seals from above. Sarah, bring the gas detector and the gap gauge.

(on the floating roof, moving along the rim) Gas detector reading 2,500 ppm at the northwest quadrant — that's well above the 500 ppm action level. And look here — the primary seal has pulled away from the shell. I can see daylight through the gap. Measuring with the gap gauge... 18 mm gap at this location. API 653 allows a maximum of 12 mm for the primary seal. This seal has failed.

The seal material is urethane foam with a fabric envelope — it's been in service for 11 years. The foam has lost its resilience — it's not pushing against the shell anymore. Mike, we have two problems: one, the seal is an emission compliance issue — we're exceeding our VOC permit limits. Two, it's a safety hazard — the gap around the rim creates a flammable vapor zone. This needs to be fixed immediately. We'll need to replace the entire primary seal — all 280 meters of it.

How long will that take and what does it cost?

The seal replacement is a 12-day job, but we need the tank empty and gas-free first. So we have to transfer the crude to other tanks — that takes about 3 days given our pumping capacity. Then 2 days to gas-free the tank with nitrogen purging. So total downtime is about 17 days. The seal material and installation: roughly $350,000. Plus lost storage capacity cost — about $40,000 per day. Total around $1 million. Not cheap, but the alternative is an EPA fine that could be 10 times that, plus the fire risk.

Julia, the insurance underwriter's survey flagged our gasoline tank farm dike capacity as a concern. We have three 50,000 m³ gasoline tanks — T-401, T-402, T-403 — all within a single dike. The NFPA 30 requirement is that the dike must contain 110% of the largest tank's capacity in the event of a catastrophic failure. That's 55,000 m³. Does our dike meet this? I need a definitive answer for the insurance report.

Let me pull up the as-built drawings. Alex, I need your help with the net capacity calculation.

The dike interior dimensions: 120 meters × 95 meters. Wall height is 4.8 meters above the interior grade. So the gross volume = 120 × 95 × 4.8 = 54,720 m³.

But we have to subtract the volume occupied by the tank foundations and the lower portions of all three tanks within the dike. Each tank is 60 meters in diameter. The foundations are 62 meters in diameter and rise 1.2 meters above the interior grade. And the tank shells themselves occupy volume up to the 4.8-meter dike height.

Let me calculate. One tank foundation volume within the dike height: the tank occupies a cylinder of 60m diameter up to the 4.8m level = π × 30² × 4.8 = 13,572 m³ per tank. For three tanks that's 40,716 m³. But the tanks are a permanent fixture, so they displace the stored liquid. Net dike capacity = gross - tank displacement + displaced foundation volume... let me be more precise.

Net available volume: 54,720 - 40,716 = 14,004 m³. But that's only for continuous space — we still need to account for the tank foundations below 4.8m. Foundation displacement per tank: π × 31² × 1.2 = 3,622 m³. Three tanks = 10,866 m³ that we don't subtract because foundations occupy space that would otherwise be available. So net = 14,004 + 10,866 (foundation volume that becomes wetted in a failure)... Actually, the correct approach: the gross containment is 54,720 m³. The three tanks occupy 40,716 m³ within the dike up to 4.8m. But in a tank failure, the failed tank would be empty at the bottom — so its displacement effectively becomes available containment. For T-401 (the largest, but they're equal), we add back: π × 30² × 4.8 = 13,572 m³. Net containment = 54,720 - 40,716 + 13,572 = 27,576 m³... wait, that can't be right either because the failed tank bottom remains. Actually, the proper NFPA 30 approach: calculate the net dike volume excluding the volume of the OTHER tanks (the surviving tanks). The failed tank's volume becomes part of the containment volume. So: 54,720 - (13,572 × 2 tanks) = 54,720 - 27,144 = 27,576 m³. This is the net volume available for containing the spill from ONE failed tank. But the requirement is 55,000 m³. We're at 27,576 m³ — only 50% of the required capacity!

That's a serious finding. You're saying our dike can only hold half of what's required?

Correct. The dike was probably designed for a much earlier version of NFPA 30, or someone used a different methodology. Regardless, we need to fix this. Options: raise the dike wall to 7.2 meters, which is a major civil project ($2M+), or build a remote impounding basin with the additional 27,500 m³ capacity. I'll work up both options for management review. Robert, in the interim, we need an administrative control — limit the maximum fill level of these tanks to 85% until the dike capacity is corrected. That reduces the spill volume to 42,500 m³ per tank, which our dike CAN handle.