Return to index: [Subject] [Thread] [Date] [Author]

Pressure Vessel Repair

[Subject Prev][Subject Next][Thread Prev][Thread Next]
Fellow Engineers,
        I have been asked to participate in discussions regarding the best way to go about repairing a pressure vessel which seems to have problems with hydrogen embrittlement.  This problem has resulted in  a major petro chemical processing plant (commercial value in the order of $50,000,000.00) being shut down.
        The vessel under consideration is 8'- 6" ID and approximately 70' tall; the material is A212 gr B with a stated tensile of 70,000 p.s.i.; the wall thickness is 3.75 inches (obviously high pressure and/or high temperature).  The vessel in inside a building about 23 feet high (hence, it protrudes through the roof by about 50').  The dry erection weight is about 300,000 pounds.  The vessel is about 44 years old.  The affected area is about five feet in length and located between 18' and 23' from the base.  The once in 30 year design wind speed is about the equivalent of 120 miles per hour.
        Some radical ideas are being considered for carrying out the repairs as you will see from the client's letter, below.  Any thoughts or comments you may wish to share would be welcome.
H. Daryl Richardson
Following is the text of the client's letter (I don't know anything more than is contained in this letter).
April 6, 2005

Here is an interesting structural problem for you. We are looking for a
simplistic assessment of stresses in their amine contactor during several
stages of repair.

They discovered some hydrogen embrittlement/laminations internal to their 3
3/8" thick contactor shell (installed in 1961) so they have depressurized
their vessel and started their summer turnaround earlier than planned. The
vessel will not be available for internal inspection until Monday/Tuesday
April 11/12. In the meantime, external NDT measures suggest that one course
of the shell is affected. The only question now is whether two quadrants of
the shell or four quadrants should be replaced. Briefly considered lifting
the vessel and sending to a shop vs lifting the vessel and repairing on site
on the ground, and current thinking is leaning towards the third thought:
lift the vessel up enough so that it is "weight  neutral" - don't lift the
vessel up off its base and anchor bolts, but use the crane for a full 3
weeks (24 hrs per day) to support the upper aspect of the vessel above the
repair, while the rotten pieces are cut out, welded in place, and stress
relieved to 1300 deg F.

The existing shell is built of A212 gr B with a stated tensile strength of
70,000 psi and a thickness of 3 3/8". I checked ASME Section VIII and I
can't find it. I checked the internet and it tells me it was replaced in
1967 with A515. I check ASME VIII for A515 and see it has a 70,000 psi grade
as well (specified minimum yield 38 ksi and specified min yield 70 ksi, 17.5
ksi max allowable stress up to 650 F and 2.5 ksi at 1000 deg F and no data
for our stress relieving temp. I'd suggest using the A515 data when
comparing calculated stresses during repair vs capabilities of the steel
plate. By the way, the repair pieces are going to be A516-70N which has the
same ASME values as this A515.

What we are looking for from you is an estimate of stresses at the repair
area during the peculiar case when the wind is blowing (plant site is in the
XXXXXXXXXXX), the crane has a hold of the top end of the vessel, the
bottom aspect of the vessel is going nowhere as it is still sitting on its
foundation (say  the nuts are pulled off the anchor bolts, but the baseplate
is still on the concrete). The attached sketch tells you where the roof line
is: essentially the entire vessel above the repair. The exact procedure for
the repair has not yet been worked out - only that stress relief would be
the last item. So the steel only sees 1300 F once all the pieces which need
to, have been replaced. I suspect that the windload on the top aspect of the
vessel during stress relieving is not an issue.

However, it has not been decided to cut out and replace one quadrant at a
time, or cut out and replace two quadrants at a time, or three. I think if
your calcs come anywhere even remotely close to a problem, we will ask you
to take a closer look at the problem.

I can't find my copy of the Alta Building Code for wind speeds/pressures in
Coleman, but I assume you have yours handy. If not, let me know.

So attached sketch shows the vessel as 3 3/8" thick, 8'6" ID, elevation of
the tan line of the vessel 69 ft above the concrete, the course to be
removed starts at elevation 18' 6" above concrete and ends at 23' 6" above
concrete. And the roofline is at 23 ft above concrete, leaving 46 ft above
the roofline exposed to the wind.

Even your prelimin thoughts Thursday (knowing you're tied up Thursday
afternoon) would be appreciated followed up with more thoughts Friday. The
crane lift would not happen until next week, but we need to have our
thoughts together before then.

Bob Rxxxxxxxx