Mono tube heat exchanger with bends

[hynatec]

I need a heat exchanger for pasteurisation of food liquid flow. The flow is viscous, i.e very low Reynolds numbers. A mono-tube solution is prefered for cleaning purposes (pig cleaning). But the pipe should be several hundred meters long to meet the final temperature. Bends could help based on the generation of Dean vortices. Are heat exchange coefficients available or should I start a CFD simulation?

 

[tickle]

In a past life I designed a heat exchanger, similar to the one you are describing. The most pratical method was to use a static mixer (took the length down to 3m or so.

The company that supplied the static mixers was Chemineer(?). I cannot lay my hands on the info immediately. They were in Derby in the UK

Obviously you cannot pig, but due to the high turbulence generated you can clean conventionally

 

[unclesyd]

We use numerous monotube heat exchangers with a very viscous material, a polymer, of various lengths. We have both the serpentine and the spiral concentric coil types.
The concentric coils have a total length of 150' of 3/4" pipe and heated by 300# PSIG steam. Process inlet 150°C outlet 210°C. On water the actual pressure drop of the coil compares well with the calculated one.

The serpentine coils are 25 12' straight sections with short radius 180° return bends. The inlet 1/3 section is 1" Sch 80 SS pipe. The middle 1/3 section is 1 1/2" SS Sch 40 pipe. The final 1/3 section is 2" Sch 40 SS pipe. The process material in the tubes is heated by vaporized oil at 295°C. The process inlet is 258°C. and outlet is 283°C.

I talk to the design engineer today about heat transfer coefficients for such a design. He stated that he had seen some coefficients for the spiral wound exchanger but not for the serpentine one. The heat transfer coefficients used were derived empirically from a single tube unit. Actually the heat transfer coefficients fall out in the third stage of the serpentine coil as it’s a 2 phase mess. All his calculations are based on pipe equivalent lengths, which I understand works quite well.

The velocity is quite low in both exchangers so we get very little help from the bends. They are in series with one intervening process step.
After cleaning the process side each coil is given an air flow check to verify cleanliness.