How to determine new capacity for heat exchanger with plugged tubes. 1

[TRBell]

How do I determine new capacity for heat exchanger with plugged tubes. Will the design "U" stay the same if over half the tubes are plugged?

I have a 31 tube, U-tube, Brown Fin tube Exchange that we have 18 tubes that are plugged with process material. We have no known tube leaks. Kerosine/olefin on the tube side and Hot Oil (Therminol) on the shell side.
Having 18 tubes plugged means that the original design:
velocity has gone up from 2.24ft/sec to 5.35 ft/sec.
pressure drop has gone up from 0.63psig to 1.4psig
effective heat transfer area has gone down from 204 ft2 to 85.7 ft2.
Original U is 57.3 BTU/(hr * ft2 * °F), but I think the U should change because I've increased the velocity (U decreases)and effectively increased the bulk shell side temperature (U increases)... but I don't know if these cancel out.

I found some correlation graphs in Rules of Thumbs for Chemical Engineers, but do not know the h for the tube and shell side, and do not know how to back calculate it from the known information...

At a minimum, just from heat transfer area, I'm probably at less than 42% of the old capacity, but I'm not sure if the additional heat now available in the bulk shell side hot oil offsets the velocity increase in the process fluid.

We did not plug the tubes with plugs, and no we're not planning on running this way forever, only long enough to get the unit back up and making what production I can while we get a new bundle built and can shut this unit down and install it. Then we can cut some plugged tubes in the original exchanger and figure out what we plugged them with. Tried a 10K blasting unit and wand is stopped 14-16 inches down tube. No process material removed at all. Normal plugging is a waxy pudding to clay consistancy, not concrete. Exchanger spec sheet and analysis, as well as tube map are attached.
Thanks,
TBell

 

[metengr]

TBell;
Go back to the heat exchanger OEM (below) and have them run a performance loss calculation based on percent plugged tubes. Most if not all OEM of heat exchanger's do this at nominal cost. The OEM can also predict temperature rise and velocity increases based on pluuged tubes.

http://www.kochheattransfer.com/cms.asp?pg=cms&cms=home

 

[TRBell]

I will contact OEM and exchanger repair shops on Monday for detailed evaluation, but I was looking for a way to figure this out on my own. Exchangers never seem to break or get fouled so you cannot get thru them on Monday morning, or mid week. They break Friday evening, or going into a 4 day weekends.

 

[metengr]

TRBell;
I fully understand because I work in Power Generation. Our heat exchangers do develop tube failures over time and require tube plugs. As a rough estimate for fluid heat exchangers, around 10% plugged tubes means you need to start thinking about either re-tube or an upgrade in replacement.

When you have the OEM or heat exchanger repair shop evaluate the loss of performance they can show you how this calculation is performed. Most use canned programs.

 

[unclesyd]

if you are required to clean the bundle again I would try to use water pressure at or above 23,000 psig with some type mechanical contrivance to handle the lance. One type . i.s shown in the link below.

I and when you pull the exchanger you might want to look at Pyrolysis cleaning to recover the bundle if the size is amenable.

http://www.nlbcorp.com/heat-exchanger-pipecleaning.asp?gclid=CKfC0Z-6wJ4CFQ8eDQod9VB1pg

 

[EdStainless]

You should be able to go back to the original design data (flows and delta Ts) and back out all of the factors. Then try to run the calcs on the new conditions. If the heat transfer was limited by the file coef on the tube side then the higher velocity may almost make up for the lost tubes. If you were shell side limited then you are hosed.
Look at your delta Ts. A basic flow/temp analysis may tell you what you really need to know.

= = = = = = = = = = = = = = = = = = = =
Plymouth Tube

 

[davefitz]

All of the above recommendations are useful.

The increase in DP from 0.63 psi to 1.4 psi implies that you yet have 63% available tubes flowing, if the reported DP is completely friction and does not include gravity head.

The convective heat transfer coeficient varies by the 0.8 power of the velocity, while the area of course varies linearly with availble surface , so the product UA varies roughly by the 1.8 power in this case . If 67% of tube flow area is availble , then you have a new UA equal to 48% of the original UA.

The best HX correction method is to use the e-NTU method of predicting heat exchanger effectiveness e. The effectiveness e is a function of the HX configuration( usually counter flow) , NTU ( = UA/W/Cp, min), where Cp, min is the flow * heat capacity of the fluid with the minimum value of this product. The mfr can perfrom this correction calc, but it seems obvious that the HX needs to be cleaned out.