Posted by Dan O'Connor on September 04, 2002 at 14:23:40:
In Reply to: Stripping Steam Controls in Crude Unit MVPC posted by Matthew Coates on September 03, 2002 at 06:51:09:
Matthew,
These are interesting questions, and ones to which the answer is, as it seems often to be in this business, it depends..
As you know, when the stripping steam is increased, the mole fraction of the hydrocarbon in the vapor is reduced.
At constant pressure, the partial pressure is reduced. However, as more and more steam is added, the vapor loading
in the tower increases, and the pressure - where the steam is being added - is not constant, it increases with the stripping steam addition.
Thus, as more and more steam is added, the increase in pressure will offset the reduction in mole fraction. In vacuum towers, as the eductors are
loaded, and adding more steam can actually raise the partial pressure of the hydrocarbon, rather than reducing it.
I'm not sure if it ever goes this far in the atmospheric tower - maybe someone can comment on this? In any case, the direction is clearly the same, and a point is reached
where adding additional steam has a small effect, and if the tower is near flooding in some section, it may be the wrong thing to do.
So the "optimum" point to run the steam depends on the hydraulic loading in the tower.
With a MVPC application - for the stripping steam on the main crude tower, I would take the stripping steam rate as an MV, and add a ratio as a CV.
This will do three things:
1.) It lets you set an acceptable range for the stripping steam ratio - usually your
process engineering support people have target ratios they like to see.
2.) You can step it during the plant testing, and measure how the material balance is changed,
along with tower loadings.
3.) If / when the tower comes up against a loading limit, you can more easily experiment with allowing the stripping
steam, versus other MV's to be adjusted at the flooding limit to see if you can determine the
"right" answer in terms of priority (profitability)
For the side strippers, I think either putting a ratio controller in the
DCS, or moving the steam directly within MVPC is an option. If the
side stripper is near flood, and has a delta p, and the flash is a critical
product spec, I would take the steam as an MV. The delta p responses should be more
linear from a steam flow rather than a ratio MV. If it is a HGO side
stripper, where it probably makes sense to simply set a steam ratio and leave it, I would
probably leave / set up a ratio controller in the DCS.
It is possible that RTO can give you the correct answer to these problems,
but it is important that "accurate" hydraulic models of the tower
and overhead fan / exchangers are included. These kinds of two phase flow pressure drop models
must be set up and parameterized properly, in order for the RTO to
make the proper trade-offs. Make sure you are convinced that your model does
model the hydraulics properly, or else it won't come up with the "optimum" answer.
If you take some good data during your MVPC testing, you can use the results of this data to compare to what
an RTO model predicts for the same changes.
The good news in all of this is, generally speaking, if the steam ratios are
set in a reasonable range, then the profit associated with changing them are not
large numbers. By moving the steam in the plant test, you should be able to measure
the steam effects - which will at least be valid for the conditions under which you've tested.
But don't wait until the plant test to find some of this stuff out -
move the steam around in your pretest, and try to determine what is the
best answer for your situation before you get into the plant test.
Good luck.