Don’t stack the deck against yourself

March 14, 2008
Always consider future expansions before finalizing layouts

A plant wanted to raise the temperature in a reactor. The new reaction conditions would push the six existing small feed/effluent heat exchangers beyond their operating temperature limit. So, the plant ordered four new exchangers, and planned to add two more later to allow removing the six original exchangers when their bundles reached their replacement point due to corrosion and erosion.

The proposed configuration for the four new exchangers had them assembled as a four pack (2 × 2) in series. The two future exchangers were going to be added on a separate foundation to make six exchangers in series (2 × 2 followed by 1 × 2). This required using space currently occupied by an air-fin exchanger that was slated for demolition as a part of a different project.

During project development the question “Are you sure that the air-fins are going to be removed?” arose as a stray thought during a review meeting. Checking revealed that the air-fin exchanger had a better than 50% probability of remaining in service. The immediate response was “Just design the new four pack to have two new shells on top.”

Redesigning the equipment and foundation to accommodate the addition of the future exchangers on the top of the stack, converting it from 2 × 2 to 2 × 3, seemed the obvious solution. However, that means the future exchangers would be inserted either into the middle of the sequence or at each end of the set of four. For heat recovery, putting the two shells in the middle is preferred. This optimizes the heat transfer by maximizing the overall exchanger corrected temperature difference. Additionally, the piping modifications are simple: the pipe segments connecting the exchanger banks are simply unbolted and re-used as-is. However, there’s a downside. Both new exchangers would handle relatively high temperatures. Putting one of the future exchangers at the cold end of the sequence reduces the mechanical design and materials severity of its service — saving up to $100,000.

However, by rethinking the layout the plant was able to double the savings. Switching to a current installation of a 1 × 1 followed by a 1 × 3 assembly allows both future exchangers to use the less expensive specification. Getting this configuration requires moving one of the existing shells from the bottom of the first stack to the bottom of the second stack and rotating it 180°. To minimize the amount of high-alloy piping the hot end of the exchangers was set as high as possible.

As this episode illustrates, even apparently obvious and easy configuration changes demand careful thought — particularly when expansion is likely. Proper pre-planning often can make future work dramatically cheaper and easier.