Name the case that sized your relief valve
A relief valve is a spring-loaded door on a vessel. Sizing it is not the hard part. The hard part is the list: every believable thing that could raise the pressure, each one sized on its own, and the biggest one wins. You never add them, because two disasters do not happen in the same second. This page is that list as a checklist, the fire case as a calculator, and one incident where the valve was the right size and was not connected.
Out now
Try Reflux Student free on Aspen Plus
Reflux Student drives your own Aspen Plus V14 from plain English. You type what you want, it opens the case, makes the change in Aspen, runs it, reads the result back and tells you what it verified. Aspen does the math. You keep the judgment. Windows, your own licence, and three free runs to start: it is a trial with a limit on it, not a free product.
Enter your email on the next page and the download link lands in your inbox, so you can open it on the Windows machine Aspen lives on. Windows will say it does not recognise the app the first time: click More info, then Run anyway. The certificate is new and Windows trusts it by reputation, which takes downloads to build. Nothing is wrong with the file.
Not on Aspen Plus? Tell me which simulator you use → and you hear the day your build exists.
A spring-loaded door, in section
The mechanism is one sentence. A spring holds a disc down onto a nozzle. When the pressure under the disc beats the spring, the disc lifts and the vessel dumps to the flare until the pressure stops rising. The set pressure is the spring. The orifice is the nozzle bore, and it is the only number the whole calculation is trying to find.
Set pressure may not be above the MAWP. A single valve on a non-fire case may accumulate to 110% of MAWP while relieving; multiple valves 116%; the fire case 121%. The relieving pressure, not the set pressure, is what the orifice equation takes.
The scenario list, as a checklist
Run this on any vessel. The point is not to find one answer, it is to be able to say out loud why the other five are smaller. Tick each one when you have either sized it or written down why it cannot happen here.
0 of 6 considered
The largest single case governs. Not the sum. A valve sized for the sum of six cases is grossly oversized for every one of them, and an oversized valve chatters: it opens, the flow through an orifice far larger than the load needs drops the pressure instantly, it slams shut, and it does that a few times a second until the seat and the guide are destroyed. The oversizing is not conservative. It is the failure.
The fire case, and a calculator
The case that usually wins is a fire nobody lit. If a pool fire burns under a vessel, the heat going into the liquid is set by the wetted surface, counted only up to 7.6 m (25 ft) above grade. Above that height the flame is not considered to reach the wall.
W = Q / L kg/h of vapour F is the environment factor, 1.0 for a bare vessel. 21,000 assumes adequate drainage and firefighting; without them the coefficient is 34,500. A is the wetted area in square feet. L is the latent heat at the relieving pressure, which falls as you approach the critical point, so a light hydrocarbon near its critical pressure gives a much larger vapour rate than its atmospheric latent heat suggests.
23.2 m2 is a 2.0 m by 6.0 m drum at half level, heads included.
Orifice from API 520 critical vapour flow at 250 psig MAWP, 110% accumulation, propane: k 1.13, M 44.1, Z 0.78, T 328 K, Kd 0.975. Change the fluid and the orifice moves; the heat input does not.
Load per case, then the orifice
Every case becomes a mass flow, every mass flow becomes an area, and the largest area picks a letter off the API 526 series. For the 2.0 m by 6.0 m propane drum above:
| Case | Relief load | Area required | API 526 |
|---|---|---|---|
| blocked outlet | 3,200 kg/h | 157.2 mm2 | F (198) |
| inlet valve fails open | 5,400 kg/h | 265.2 mm2 | G (325) |
| tube rupture | 4,100 kg/h | 201.4 mm2 | G (325) |
| thermal expansion | 114 L/h | 0.5 mm2 | D (71) |
| utility failure | 6,200 kg/h | 304.5 mm2 | G (325) |
| pool fire | 9,523 kg/h | 467.7 mm2 | H (506) |
the sum of all six 1,397 mm2 → L (1,841 mm2) Adding them buys 3.6 times the orifice area, and a valve that chatters on every case it was bought to protect against.
The valve that was the right size
A correctly sized relief valve only works if it is connected. In 2013 a plant added two block valves between a reboiler and its relief valve so the reboiler could be cleaned online, and left them shut. There was then no path from the shell to the valve.
Two block valves are added between the reboiler and its relief valve, so that a fouled reboiler can be isolated and cleaned while the unit keeps running.
The reboiler is standing idle and isolated, with propane trapped on the shell side and both block valves shut. Neither valve is on any check that would have caught it.
Hot water is admitted to the tubes. The trapped propane boils. The pressure has nowhere to go, because the only relief path is behind two shut gate valves.
The shell ruptures and the released propane ignites.
Two people are killed and 167 are injured. The relief valve was correctly sized for this vessel. It was simply not connected to it.
The lesson is not about sizing. It is that a relief path is a system, and every block valve in that path is a way to delete the valve without removing it. If a block valve has to exist, it is car-sealed or locked open and it is on a checklist somebody signs.
The simulator does the flash. It will not do the listing.
Any process simulator will flash a stream at 110% of MAWP and tell you the vapour rate to four decimal places. Not one of them will tell you that a tube rupture is credible on this exchanger, or that the line you just blocked in is liquid full. That judgment is the work, and it is the layer we are building at Reflux: the model does the arithmetic, you keep the listing.
Try Reflux Student free on Aspen Plus →
Built from a reel. If something here is wrong, tell me and I will fix the page. ← reflux.sh