Reqirement for symetrical piping on the inlet of a CDU heater 2

[Tyal51]

Has anyone converted the standard inlet manifold on a crude heater to a symetrical piping layout to eliminate uneven temperatures in the passes?

 

[EmmanuelTop]

Why not apply control valves for each pass?

Another option is to use restriction orifices in case that piping is not symmetrical. Not sure how actually symmetric one can make it in reality.

http://antwrp.gsfc.nasa.gov/apod/astropix.html

 

[Tyal51]

Emmanuel,

The existing piping configuration does not lend well to the installation of control valves. We are looking at replacing the entire inlet manifold with a symetrical configuration.

Wanted to know if anyone has done this on a small process heater (20,000 BPSD crude).

Regards

 

[owg]

I would think restriction orifices would be much cheaper than replacing the manifold. Perhaps the manifold is being replaced for additional reasons.

HAZOP at

http://www.curryhydrocarbons.ca

 

[EmmanuelTop]

I agree with OWG. If flow distribution is the only reason for replacing the entire manifold, installing orifices is much cheaper and quite effective solution. It doesn't matter whether the heater is bigger or smaller size one - hydraulics work in both cases.

http://antwrp.gsfc.nasa.gov/apod/astropix.html

 

[Tyal51]

Thanks for the replies - appreciated

 

[popovic]

Crude heater inlet manifold deals with the liquid phase and a standard design is ok, assuming control valves or even orifices are balancing coil flows. Crude unit heater service is not a typical coking service and can tolerate some coil flow variations +/- 5 %, while focusing on firing balancing. If replacement is taking place for other reasons the most symmetrical arrangement is always preferred.
We have a symmetrical inlet manifold (bifurcated 1/2/4) arrangement all in a horizontal plane but in the 4 pass NHT heater to handle potential flow distribution in two-phase flow without flow control valves. A local PI at coil inlet roughly indicates flow/flow distribution. This design doesn't work so well at low turndown rate/flows.

 

[Tyal51]

Hello popovic,

The existing piping set up is fitted with 4 independent lines from a common header. Each leg is fitted with an orifice plate and a globe valve - to allow for "fine ture" balancing in each coil.

What we have deduced is that we have excessive two phase flow - 8wt% causing excessive vaporization with a calculated velocity in the vertical lines of ~30 ft/s. The original design was for 3wt%. In order to balance the heat in each coil (measured inside heater) first leg needs to be almost closed and we suspect that this will cause coking of the tubes.

Our conclusion is that we need to split the header into 1/2/4 configuration to get better distribution.

Thanks of the input.

 

[today8]

Hello Tyal51,
it seems you have a significant pass flow imbalance with first leg coking risk. Pay attention to equal firebox firing pattern to prevent local overheating of this coil area. Relax operating criteria for coil flows variation from +/-5 % toward 10 % to fairly equalize outlet temperatures. Check for coil vibration and adjust process/fire side accordingly.
If you are going for a symmetrical inlet lineup I would suggest few guidelines: design inlet manifold bifurcation from 1 to 2 to 4 passes in one plane. Follow gradual diameter reduction up to the heater inlet flange. Layout of the inlet manifold such that there is a length of min. 10 ODs before the next split is made. (Splits must be in one plane). Install PI at each coil inlet close to the heater inlet flange.
Provide flanged spool at each coil close to heater inlet for decoking purposes. Review the heater start up procedure concerning pass flow stalling.

 

[popovic]

Sorry, on last one should be "popovic" instead of "today8"