Apr-2012

Column preheat improvement to debottleneck furnace operation

Simulation helped to identify and mitigate the root cause of a furnace bottleneck to maximise the removal of light ends in the feed to a preflash column

Said Al Zahrani, Samit Roy and Edwin Bright
Saudi Aramco

Article Summary

A preflash column’s primary purpose is to reduce the load on the downstream furnace. Refineries use a preflash column with the preheat train network to maximise the removal of light ends in the feed so that the furnace inlet temperature can be raised with the available hot streams in the refinery. Light ends need to be removed to prevent vapourisation in the furnace inlet. In this article, we explain how the root cause of a furnace bottleneck was identified and mitigated.

In the plant, a high stack temperature in furnace F200 (see Figure 1) was found to be limiting throughput. Operating the unit with design throughput with the furnace limitation would cause the loss of diesel in the C200 column bottoms, which is atmospheric gas oil (AGO). It was found that the change in feed quality and fouling were the primary reasons for the furnace bottleneck, and this has been addressed by adding three additional shells in the preheat train.

Process description
The preheat exchanger network of a Saudi Aramco refinery processing condensate is shown in Figure 1. The network contains both preflash drum D100 and preflash column C100. The preflash drum vapours go to the preflash column and the preflash column side cut-medium naphtha goes to the 53rd tray of the condensate fractionating column C200, which contains 58 trays. The overhead of C100 and C200 goes to the overhead condenser and reflux drum (not shown in the figure). This reflux drum provides reflux to C200 only. For C100, the reflux is provided by preflash column pumparound.

Stabilised condensate from storage is preheated in a series of exchangers (E355, E212, E300 and E101) before it reaches the desalter, D100. The condensate after the desalter is further preheated in another series of exchangers (E231, E222, E211, E241, E224, E201, E221, E233, E230, E240, E243 and E200) before entering the preflash column (C100) to remove the light ends.

The bottom liquid from the preflash column is routed as feed to the condensate fractionator C200 through furnace F200 after exchanging heat in exchanger E244, AGO recycle, dirty wash oil recycle to obtain the desired flash zone temperature at C200.

In the condensate fractionator column C200, the desired products are separated via three pumparounds and four side strippers. The side stripper products are heavy naphtha, kerosene, light diesel oil and heavy diesel oil. Atmospheric gas oil is withdrawn from the bottom. Top unstabilised light naphtha from the C100/C200 overheads is routed to stabiliser C350. Here, light naphtha and LPG are separated.

Methodology
• A validated process simulation model was developed for the heat exchangers and columns to reflect current operation with regard to flow, pressure and temperature profile
• Extra area has been added in the exchanger where possible to increase the preheat temperature to C100
• While simulating C100 for the target case, medium naphtha has been kept at 5°C less than the platformer feed end point
• The new temperature profiles in C100 and C200, along with simulated product draw-offs, pumparound flows and duties, were used as the exchanger inputs for determining the C100 preheat temperature
• For this preheat temperature, a reduction in the required furnace duty was calculated.

Results and analysis
Simulation of exchangers and fouling analysis
The performance of each exchanger in the preheat train network has been evaluated using Aspen EDR. The fouling coefficients in the condensate side have been adjusted to match the performance of the outlet temperature of the exchangers. The analysis indicated that 11 out of the 16 exchangers are at the fouled condition (observed fouling more than the design fouling). See Table 1 for the design and observed overall heat transfer coefficient of each exchanger.

Fouling is found to be severe in E300 (service: fractionator overhead/cold condensate). The estimated overall heat transfer coefficient is only 21 Btu-ft/hr sqft°F compared to the fouled overall heat transfer coefficient of 71.9 Btu-ft/hr sqft°F. (Clean overall heat transfer coefficient is 99.1 Btu-ft/hr sqft°F.) In other words, the preheat temperature would have been increased from 425°F to 430°F with design fouling in E300. The fouling could also be due to salt deposition in the fractionator overhead circuit.

The cause of fouling is not asphalt precipitation for the following reasons:
• Condensate is a good solvent for asphaltenes
• Fouling has also been observed in the low-temperature exchangers (less than 250°C). Asphaltene precipitation begins to be a problem above 250°C only.

Reasons for furnace bottleneck
The C100 preheat design temperature is 473°F, and that achieved in the plant was 425°F. Hence, a case study was performed on a validated heat exchanger network with design fouling coefficients. The C100 preheat temperature improved from 425°F to 443°F. Next, the design product flow rates and draw temperatures were used to determine the C100 preheat temperature. The obtained C100 preheat temperature was 472.3°F, which is very close to the design value of 473°F. The contribution of fouling in the decrease in preheat is around 37% and that is because the change in feed quality is 63%.

Now let us analyse the effect of a decrease in preheat temperature on the furnace load. Table 2 and Figures 2 and 3 illustrate the discussion.

The absorbed furnace duty calculated from the actual case simulation is 539 MMKcal/hr. The actual heat release is more than the design because of lower efficiency and a higher duty requirement. The higher duty requirement is due to the change in condensate feed quality and fouling in the preheat exchanger.