PTQ Q3 2025 Issue

4500

800,000

Upper reboiler duty relatively constant until tray runs dry.

4000

700,000

3500

600,000

3000

500,000

2500

400,000

2000

300,000

1500

200,000

Lower r eboiler heat input leading up to relief is more than 50% of normal.

1000

100,000

500

Time (min)

Pressure (kPag)

Relief load (kg/hr)

Lower reboiler duty (MW)

Upper reboiler duty (MW)

Figure 4 Alky fractionator TPF dynamic relief results

input estimated based on detailed calculation of radiative heat transfer from refractory to tubes. The calculation was built into the dynamic model and updated with each time step. Results of the dynamic simulation of this scenario are shown in Figure 4 . The dynamic relief load was found to be 684,000 kg/hr, about 55% higher than the UBH relief load. This increase can be attributed to the following factors: • Residual duty in the lower reboiler was more than 50% of normal duty during the period leading up to relief, signif- icantly higher than the 25% assumed. • Upper reboiler duty remained relatively constant for the duration of relief rather than falling due to reduced LMTD. This dynamic model demonstrated that key assumptions incorporated in the UBH model were not valid for this par- ticular system: • Due to a particularly rapid relief event (approximately five minutes between the initiating event and peak relief), resid- ual heat input from the lower reboiler (fired heater) was cal - culated at more than 50% of normal duty during the period leading up to relief, which is significantly higher than the 25% assumed in the UBH calculation. The detailed thermal model predicts that the residual heat input will decay rap- idly within the first 10 minutes, demonstrating that 25% is likely conservative for most systems, which will take more time to reach relief pressure. The time to reach relief pressure is a function of the relief load, the system volume, and the margin between operating pressure and PSV set pressure. • Due to the reduction in heavy vapour rising from the bottom of the column to the upper reboiler feed tray, the hydrocarbon feed to the reboiler became lighter over time. This effect was not anticipated by the engineer perform- ing the UBH calculation. As a result, the credit taken in the UBH calculation for reduced LMTD (based on constant composition) was not valid. The process temperature did

not rise significantly, and duty remained relatively constant throughout the relief event. In most systems with a single reboiler at the bottom of the column, the assumption of constant composition in a TPF scenario is conservative with respect to LMTD, as the bottoms material will get heavier and hotter over time. Case study 3: methanol distillation column loss of condenser The initial relief load analysis for a methanol/water distilla- tion column was performed using the UBH approach (see

Relief

Cooling water

Overhead condenser

FIC

Overhead drum

Methanol

Overhead pumps

60% MeOH feed

Methanol distillation column

FIC

LP steam

Condensate

Reboiler

Water

Bottoms pumps

Figure 5 Methanol distillation column system

PTQ Q3 2025