3 replies to this topic

[obsidian_dust]

Hi everybody,
I would like to share my experience regarding Crude Distillation Unit Air-Cooled Overhead Condenser (later will be simply called as Condenser) in our Refinery. We have two CDUs and each of them is equipped with a number of banks of condenser. Each CDU (including the condensers) was designed by different vendor, so there is a difference in their manifold piping configuration.
The first CDU has this kind of manifold arrangement:



And the second CDU has this kind of manifold arrangement:



After years, we saw the significant difference between those two condensers in regards to their outlet temperature profile. For the first CDU condenser, we observe that there is a big variation in outlet temperature between each condenser bank. The coolest outlet temperature could be around 35 deg C while the hottest one could be more than 70 deg C (the inlet temperature is around 160 deg C). It means that the difference between the hottest and the coolest outlet temperature is about 40 deg C. Meanwhile, for the second CDU condenser the difference between the hottest and the coolest outlet temperature is only about 10 deg C. All fans work well and there is no significant difference in performance between fans in a same condenser.
Once I read a design practice from a process licensor mentioned that for one phase fluid (overhead of a column is considered as one phase fluid), the preferred/recommended manifold configuration is like the one we use in the second CDU.
My question is, does this manifold arrangement significantly affect the CDU condenser performance in providing proper flow distribution?
If somebody has any relevant idea, theory, or experience, please feel free to share.
Best regards and thanks in advance.

[Zauberberg]

The second arrangement is much more practical and efficient, as it eliminates flow distribution issues between the header and sub-headers, as in the first arrangement.

The key thing is that the inlet/outlet headers should be sized big enough in order to eliminate any substantial pressure drop along the header. Then, having both headers operating at essentially constant pressure (P1 for inlet, and P2<P1 for outlet header), the driving force for fluid flow through each individual cooler will be identical, and it will result in almost equal flows for all exchangers. In reality, there are some deviations of course.

As long as you mantain the performance of each fan close to each other, coolers will be performing similarly as well. Once when you start experiencing fouling in exchangers, the performance of indivivual coolers will start to vary depending on the extent of fouling in each exchanger. Coolers that are more fouled will show higher resistance to flow, and that will push more flow through other exchangers. You will observe that by tracking outlet temperatures from the exchangers - less flow resulting in lower outlet temperature.

[obsidian_dust]

Zauberberg,

Thank you very much for your helpful answer.

Best regards.

[Zauberberg]

Please note that this rule of arrangement applies only for single-phase flow at the cooler/condenser inlet (and assuming inlet header running liquid-full in case of a liquid cooler). If there is two-phase flow at the condenser inlet, this calls for special design.

Having everything else equal, in case of two-phase flow, liquid will tend to flow towards the farthest exchanger due to inertial forces while gas/vapor will flow through the nearest ones. This would, just by itself, create an imbalanced heat transfer between the exchangers.