HOW TO DO AN INSPECTION OF A DOUBLE WALL CLAD REACTOR 1

[lganga]

The Vessel is a 70’ x 10´ double wall reactor. Inner Shell is a 30 mm thickness wall CS with 316L 3 mm clad.
Outer shell is a CS and it contains the cooling or heating water, depending of the stage of the inner vessel reaction.

We want to check the thickness of the inner wall with clad steel but the only way to do it is from the inside of the vessel.

As we do not have any Ultrasonic thickness device that measure the clad and the CS from the clad side of the vessel (inside inner shell).

Question: Have anyone of you found this kind of problem before?. How did you solve it?
Thanks

 

[unclesyd]

Here is what I've done several times. You are going to have to cut a few inspection ports in the outside shell to have access to the inner shell for thickness. I would look at cutting and threading of the holes to receive a screw plug. The hole will depend on the clearance between the two jackets. Depending on your water and steam quality one or two ports may work or it could take several more. When you cot out the outer shell you can verify your thickness measurements of the outer Shell
Measurement of the clad is straight forward.
I
would recommend that you do UT of as many of the connections between the two shells. It will take a highly skilled UT technician to accomplish this. I think this is important because of the heating and cooling cycle. I have seen a lot of cracking at both the heal and toe of the welds.

 

[lganga]

unclesyd

you sent me a good procedure.
IN that way i should have an spot picture of both sides.
But for the connections some part of the UT should be done inside, that means from the clad side, not any UT will be give us the real data.

I thought that some UT technique was used for checking the clad and the steel at the same time. This kind of problem are in many vessels, also in the nuclear power plants

I will send some details later.

I heard that some of this type of polymerization reactors have had cracks on the joints between the outer shell and the header of the inner in the upper closure plate,the crack went into the header and the header got broken

 

[unclesyd]

Here is a tentative procedure.
If you have access to the clad for the inside do a PT to ascertain if the clad has been compromised. If you find a hole in the clad make sure you open the outer shell in this area to expose that area for UT testing. If repair are required on he clad think about removing a section and then get UT measurements of the inner shell from the inside. If repairs are not required then you may have to go the outside route. Depending on how the clad is attached you also may be able to open small areas for UT from the inside Determine clad thickness.
Depending on the results from the inside go with the access ports. You may have to get a long reach UT transducer to facilitate the inspection.

You are going to have to be flexible on your inspection, in other word change it for any workarounds needed.

More Information would be a great help and get you more suggestions.

 

[lganga]

unclesyd (Materials).

Here attached is the file where you can see a drawing of the upper header and to the right side , is a detail of the double wall and the closure of inner header and the external wall.

This can give you an idea of the vessel and the access to different points.

I was thinking to weld some 2" #3000 screwed niples as inspection points of the inner wall.

Regards

 

[unclesyd]

That is initially what I had in mind using a half coupling, but after looking at the print it may prudent to add a larger access port or a hand hole.
There are some troublesome areas that are going to be extremely difficult to inspect. These include but not limited the large nozzles which may be able to be checked from ID of the nozzle, The closure welds on the jacket, and the welds around the stays.
By positioning the access port in the right place you could at least use a video probe to look at some of the welds.

I'm having trouble with the drawing on my little screen and even when I print it's still a little blurred. These conditions will help in deciding what route to take.

Before we go on, could you post the design conditions of the vessel.

http://www.laboiler.com/Products/Handholes/handholes.htm

 

[r6155]

Try with Acoustic Emission Examination -AE-(see ASME V)to verify structural integrity.
Consult your local AE company

Regards

r6155

 

[Duwe6]

AE is very good with identifying and locating cracking, terrible with wall thickness. The 'gold standard' for the inspections you need is to hire an API-510 Inspector and give him/her internal access. Possible problems with the cladding are typically identifiable with a knowledgeable, close visual inspection. These areas can be tested with straignt-beam UT using a FLAW DETECTOR. A digital thickness meter and/or dual-element transducers are about useless to 'see' the bond/disbond status to the cladding. If the bond is good, the total thickness of your vessel can be measured at that point. No holes needed in the jacket.

 

[lganga]

r6155 (Mechanical)
We did AE 10 years ago with the company Physical Accoustic, but if you see the vessel that I have attached in this site, you can realize that the information you get is only from the outer wall, because you can only put sensors in the outer wall. Even the information obtained is not very confident. there are a lot of attached internals, as water deflectors, welded, that gives a lot of noise and you are not able to find exactly where the noise is coming from.
Then you do not get any information of the inner wall reactor , where the reaction is on and pressure is higher. Outer wall is easy to inspect with UT thickness or other techniques.
so I am looking for alternatives
Regards

 

[unclesyd]

I agree with you about AE as I was in on the testing of the MonPac system. There are two of these systems in the basement of our lab. We tried to look at our jacketed Autoclaves and got nothing but hash and some overwork recording oscilloscopes. Our biggest sucess with AE was our Horton Spheres (5) and some rail cars.

A point was brought up in an above post as to the nature of your cladding, is it mono, strip, plate, or welded?

As posted before please comeback with you design conditions.

Some of the welds you will have to do with a partial visual with a video inspection scope. This is not all that bad as cracking in CS tends to open up under cyclic service and normally starts on the outside wall and at the welds,

 

[lganga]

unclesyd (Materials9

About AE:
I am glad to read your comments about AE. We did the spheres and the horizontal vessels wiht AE.
As a comment I should say is that the triangulation of the sensors is very important, and no many vendors take care about that point .Is a simple and tedious job and no many are aware of that problem. specially when you are working with a 20 meters diameter sphere.

About the vessel:
Design conditions: 18 bar/8 (inner side/outer)
Operationg conditions: 15/6
the Shell is a cladded plate, with each course of 2060 mm wide.
I have attached an scanned drawing with more image definition. I hope you can see it
Regards

 

[unclesyd]

I talked with some old colleges about your problem and they essentially agreed with what I posted. If you use pipe nipples use a half nipple if possible. If possible use a small video probe with recording capabilities.
One of the more important points that I failed to emphasize is to have a tentative plan and get with your AIA. They are always more receptive when you don't surprise them. Be sure to discus the number of inspection ports required and try to reduce the number if no anomalies are found on the first two or three, using 12 as the minimum number for a non-jacketed vessel. The head and nozzles should be straightforward. Ask about adding handholes as you have liquid in the jacket.
If you can locate the t junctions on the inner shell I would try a radiograph or two. You want get the detail but will see anything detrimental to operation that requires further investigation. Film on the insde.