Dear Sir,

We have a heat fixed tubesheet exchanger having following data :
Shell side design pressure : 175 Bar
Shell side design temp. : 120 C
Tube side design pressure : 12 Bar
Tube side design temp. : 100 C
Shell / Channel Material SA 516-70
Tube Material SA 213-TP 304
Tubesheet Material SA 182-F304
Shell Side Fluid : AMONIA
Tube Side Fluid: WATER
Service: LETHAL ( Being amonia in shell side)
Shell to Tubesheet joint: Butt joint (100 % RT) Refer : UW13.3 (a)
Tube to tubesheet Joint: Strength Weld with light rolling (As per UW-20.1 (C))

After tube insertion, tubes were 3 mm projected out side of tube sheet face. Prior PWHT tubes wer kept free, now after PWHT of shell, some tubes gone shorten in length about 8 mm and are remained in side the tube hole from face of tube sheet. Now for tube to tubesheet joint, can we go for to build that tube to increase the required length of 8mm ? If YES than please suggest procedure.

Dear Jugno Lal,

please specify what do you mean with "to build that tube to increase..."

Best regards

Michael

Jugno,
Why don't you provide some more info.
1. Why PWHT was done before tube-to tubesheet welding but after tube cutting and fitting
2. What type&technique of tube-to-tubesheet joint was planned to be used and what other possibilities are there to choose from?
3. Give diameter/wall thck/pitch of your tubes
4. Describe whether there is some access to the tubes so they can be removed
5. What does it mean that SOME tubes got shorted? Unequal material props? Bulged tubesheet? Bent shell?

_________________
Konrad Anikiel

Dear Mr. Michael,

This excahnger is fixed tube sheet design (NEN Type) Shell material is CS and Tube & tube shhet are of SS material. Bundle was inserted in shell and tubes were kept free with out any tack weld or proper strength weld. After bundle insertion and completion of welding of shell with tube sheet, PWHT was carried out. After PWHT some tubes was shorten in lentgh and remained beyond the tubesheet outer face. The reduction in length of tube is about 8 mm. So "To build the tube to inrease the length" means to increase/make length of tube by welding electrode.

Dear Mr. Konrad Anikiel

Here is requisite info:.
1. Tube were of exact length. PWHT was done as this was requirement of shell side.
2. Planned method is manual.
3. Tube dia is 19.05 mm, thk= 2.77mm, pitch 28.57mm
4. Channel is still not welded to shell, so tubes are in full access, but we cant replace the tube due to unavailability.
5. All tube lengths were proper. Tubes became shorten may be of reason of bulgging or any other reason. Refer above answer to Mr. Michael's querry too.

UW-20 has changed a lot in current addenda, in case you didn't notice. There are no sketches anymore. Means more freedom to the designer.

Although Appendix A doesn't prohibit from such a long buildup, I wouldn't do it.

Look at Sketch 5 Fig A-2. The offset dimension, marked "0 to t max", if made bigger (it's nonmandatory appendix!) might probably be acceptable, but I would discougage you from it as well.

My suggested solution is to make a countersunk recession around the tube hole:

With your dimensions you should be able to do it.

But I would advise to put some efforts to find the reason why not all of your tubes have this problem. Maybe they were not properly supported by baffles and got drooped inside the shell? Can you see thru them?

_________________
Konrad Anikiel

Hello Konrad,

Figure A-2 sketch (5) is a good advice. When the tubes were 3 mm outside the tubesheet surface prior to PWHT and are shorten for 8 mm, this means they are 5 mm inside. With a wall thickness of 2.77 it´s a possibility to go, at least theoretically. However, I agree with you that this is difficult to manufacture.

On the other hand your suggestion is not simple either. Moreover you have to run a new calculation for your heat exchanger when you reduce your tubesheet thickness that way. 3 mm less could cause new troubles. Nevertheless, most important is your question where the missing 8 mm are gone.

Besides: Why there are no sketches in UW-20 in your code book anymore? Have you removed page 130.1?

Best regards

Michael

Well, I don't fully understand whether it's 8mm below the tubesheet both sides, or one side, or +3-8=-5mm, or what. Ant what "some tubes" means- just few spots here and there or 75% of all tubes, for instance. And it's not clear whether it's TIG or electrode welding.

Right, I removed 130.1- thanks for pointing. And UW-20 doesn't have such sketch, but it doesn't mean that it's prohibited.

I don't think a few local recessions would weaken the tubesheet significantly, but again- it depends on how much is "few"

_________________
Konrad Anikiel

Dear Mr. Michael & Konrad Anikiel

We can not go for sketch 5 fig A-2, because of tubesheet thickness reduction. And in this way, fillet weld size will be equal to tube thickness 2.77 mm, whereas for strength weld required fillet size is 3.25 mm, that's why, we have selected sketch 3 fig A-2. Some tubes means about 25 % and this reduction in length is from both sides. Your understanding of length reduction is very right.

Actually we have three exchangers of this type with very little difference in design data. In two exchangers, tube to tube sheet joint was welded prior PWHT. So after PWHT, there is no problem. We got this problem in only one exchanger in which tubes were kept free. So different linear expansion of CS and SS materilas might be the reason of such reduction in tube length.

We are considering for to build up that reduction of 8 mm to take these tubes on its required/actual position as looks like in Sketch 3, Fig A-2. Now question is that, can we proceed in this way and does code allows us to do so or any good engineering practice allow for this?

Regards,

Jugno

Dear Jugno,

I´m afraid you cannot convince your AI that a part of your weld becomes tube, and another part stays weld. Please note that the sense of the calculation of the length a is to determine the contact area between tube and tubesheet. You don´t have sketch (3), therefore you cannot use the "imaginary" length a to fulfill your required length of 3.25 mm.

However, when I got along with you correctly, the result of your welding will look like sketch (7). Why don´t use it? Moreover this will be the only way to satisfy the Code requirements regarding a, isn´t it?

Michael

Jugno, difference of thermal expansion between shell and tubes is bidirectional during heating&cooling. It has nothing to do with your shortened tubes. And it's the same for all tubes, not just for some. I'm afraid, it's not true that the tubes got shortened. Perhaps it's the shell that got elonged. This might induce high stress in the tubes that were welded before PWHT in two remaining heat exchangers. Just imagine how much stress you have to induce to get such a tube elonged by 8mm! My advice is to measue the shell and check this. High residual stress will probably cause pretty quick corrosion in tube-to-tubesheet welds. Remember that this is lethal service and you cannot neglect such threat.

_________________
Konrad Anikiel

For me it's curious who designed such a process, where a lethal service big heat exchanger has one pressure almost 15 times bigger than the other. What about a 10 to 13 rule?

Jugno: Consider removing the tubes that got short, dividing them somewhere in half and orbit-welding a short piece in. Just find some excess pieces in scrap heap.

_________________
Konrad Anikiel

Dear Konrad Anikiel,

You are suggesting for joint in tube. For lethal service and for SS tube, is joint allowed ?

Dear Michael,
you are suggesting for using of Sketch 7 Fig A-2. YES this seams something acceptable. Will you please suggest weld calculations for that particular joint. After having these calculations, I will take up this matter with AI.

Regards,

jugno

Hi, folks. Are you wondering why tubs are shortened?
1st CASE: Both tubes and shell same material: When PWHT to complete HE is made, great attention shall be given to have same temperature in shell and tubes. This is accomplshed by placing a fan in the oven which blows air through the tubes or keeping a very very low heating and cooling rate. If temperature is not uniform tubes will bulge. When tubes are welded one side only and tube are not completely free o move because of bad alignment of baffles they will bulge and one side get shortened.
2nd CASE: Different material and different expansion. This is our case. Lethal service requires PWHT. Good to weld one side only and perform local PWHT on tubesheet where tubes were welded after the first PWHT in oven. This is the correct way but tubes were not free to expand because of too much friction in baffles and bulged, remaining short. Bad practice was followed when welded both sides before complete PWHT: a simple calculation about the difference in elongation during PWHT to know how much compression was made during heating and tension during cooling. But...... which was the temperature within the tubes during PWHT?
In these cases it is the weld tube to tubesheet to soffer: PT on fillet welds of tubes to tubesheet will ensure no damage occured.

Sorry I did not take note of letters before so I do not remember which sketches you are going to use. Sketch in Fig A-2(5) seems appropriate to your case but gap between pipe end and tubesheet face shall be limited to 2xthk of the tube (1 x thk from 45° fillet edge). Build-up of tube end sounds very unusual and needs a lot of courage to develop, although there is no prohibition in the Code unless the "good engineering practice". If applied mind full PT and a qualification test by pull-throgh as recommended in Appx.A, and good also would be a qualification test as required in Div.2 (sectioning) would be recommended.

Another point to consider is: Is it required to PWHT tubes to tubesheet welds in L service? or it is exempted per Fig UCS-56 (for PNo.1) Note (2)(c)(2)? At this question Art Spencer (FM Manager and Honorary Member in ASME Committee) years ago answered that there is no exemption: Par. UW-2(a) statement rules above Fig. UCS-56 and PWHT is required. But I am still not fully convinced.
Mauro.

Jugno: can you tell whether you can see thru all the tubes in all your heat exchangers so we can verify our hypotheses on bulging? Otherwise it's all close to chiromancy. My concern is whether you have high residual stresses in welded tubes beacause that might be much bigger problem to you than the shortened tubes. Can you measue length of the shell after PWHT and compare to before PWHT?

Mauro: what do you think about my fears? Am I crazy?
Konrad

_________________
Konrad Anikiel