You are on page 1of 11

josedipof (Mechanical)

2 Sep 12 4:18

Gent,
general question about test pressure according to ASME VIII div 1,
Hydrostatic pressure=1.3xPdxStest/Sdes (UG-99(b))
Static pressure can be neglect (very low)
Assume MAWP=Pd
Min. MAWP calculated < Hydrostatic pressure
Min. MAP calculated < Hydrostatic pressure also
In this case do I need to check the stress in the vessel at hydrostatic pressure(with 90% of the
yield) or is implicit in the code that it will be always verify?
brkmech1234 (Mechanical)

2 Sep 12 5:20

Please restate your question which seems to be not clear


austsa (Mechanical)

2 Sep 12 7:48

Sress due to hydrostatic pressure will always be less than yield.


josedipof (Mechanical)

2 Sep 12
14:10

Sorry if it was not clear, i try it with an example.


1)
Pd=50 bar
1.3xPdxStest/Sdes=65 bar=Ptest
MAWP of the vessel=53 bar
MAP=55bar
Is the vessel safe at P=Ptest? If yes shall I prove that and how?
2)
Now the test pressure is according to PED
Ptest=1.43Pd=71.5 bar
Same question as above
brkmech1234 (Mechanical)

3 Sep 12 0:39

In no case the stress due to hydro test shall not execeed the yield stress. Good practice is to
keep it below 0.9 times yield stress. This check is mandatory as per the code. However the
below is worth noting
Generally test pressure is 1.3 times times the design pressure. But some times it is as high as
1.5 times of design pressure/MAWP. Most of the cases it is Ok because, the allowable stress of
material according ASME VIII Div1 is min(Tensile strength/3.5, yied stress/1.5). Thus when you
back calculate the stress at hydrotest pressure as high as 1.5 times MAWP, it will not excced
the yield stress. More over, yield stress at design temperature will be lesser than that at room
temperature where the hydro test is carried out. Thus, the stress due to hydro will be less
than 0.9 times yield. Howver, in few cases it may exceed yiled where the static head is high
like columns.
josedipof (Mechanical)

3 Sep 12 2:01

brkmech1234 thank for your reply


only one remark, where is state in the code that the check (hydrotests with 90% yield) is
mandatory?
brkmech1234 (Mechanical)

3 Sep 12 4:10

It is mentioned somewhere in UG 99 . Please check


josedipof (Mechanical)

6 Sep 12 2:15

that is the point, nowhere is mentioned that it is mandatory!

Confidential 2012 Aker Solutions

Page 1 of 11

brkmech1234 (Mechanical)

6 Sep 12 3:31

josedipof
Do not jump to conclusion. Please refer to last sentance of UG 99(b)
"All loadings that may exist during this test shall be given
consideration."
What do you read out of it. I read it like one should check the integrity of the vessel during the
test. It is simple fact that any material stressed beyond yield point has essentially to fail.
TGS4 (Mechanical)

10 Sep 12
15:04

Please keep in mind that these limits on the state of stress during hydrostatic test are ONLY
applicable to the general membrane stress - that is it does NOT include any local stresses.
It is permitted, during operation for some localized stresses to exceed yield - even by a factor
of two. Therefore, it is logical that some localized stresses will exceed yield during the
hydrostatic test. This is OK, though.
Please do not use generalizations like
Quote (brkmech1234)

In no case the stress due to hydro test shall not (sic) exceed the yield stress

Although that may be true for the case of general membrane stress, it is not true for all other
situations.
Furthermore, your statement of
Quote (brkmech1234)

It is simple fact that any material stressed beyond yield point has essentially to fail.

No, it is not. If it has exceeded the ultimate strength in uniaxial loading, then it is considered
to have failed. For many materials, signifcant strength improvement is obtained from
exceeding the yield strength: strain hardening. In fact, there are some who argue that the
strain hardening and the subsequent compressive residual stress in local areas is a primary
function of hydrostatic tests; it's much more than a leak test.
brkmech1234 (Mechanical)

11 Sep 12
23:54

TGS4,
Your comments seems to be bit clumsy. Let me have some time to understand and comment
on them.
I have a question. You are aware of the context which I put the above comments. Do you
agree with Josedipof's statement that check for hydrotest is not mandatory?
EngAddict (Mechanical)

12 Sep 12
3:20

"where is state in the code that the check (hydrotests with 90% yield) is mandatory?"
UG-22 j)
TGS4 (Mechanical)

13 Sep 12
14:00

To answer the OP's question - are there any additional checks to be made when calculating
the hydrotest pressure - according to UG-99, the answer is YES, but not of the type that you
indicate.

Confidential 2012 Aker Solutions

Page 2 of 11

One of the most important parts of calculating the hydrotest pressure is not the multiplication
of the design pressure by 1.3 (for ASME VIII-1), but that this product is also multiplied by the
lowest stress ratio (LSR) for the materials of which the vessel is constructed. The stress ratio
for each material is the allowable stress value at the test temperature divided by the
allowable stress at design temperature. Bolting is NOT included EXCEPT when
1.3xLSRxS_bolt(@ test temperature) > 0.9xSy.
Note article UG-99(d)
Quote (UG-99(d))

The requirements of (b) above represent the minimum standard hydrostatic test pressure required by
this Division. The requirements of (c) above represent a special test based on calculations. Any
intermediate value of pressure may be used. This Division does not specify an upper limit for hydrostatic
test pressure. However, if the hydrostatic test pressure is allowed to exceed, either intentionally or
accidentally, the value determined as prescribed in (c) above to the degree that the vessel is subjected
to visible permanent distortion, the Inspector shall reserve the right to reject the vessel.

So, in Division 1 there is no longer a maximum based on general membrane being less than
0.9xSy. The limiting criteria is "visible permanent distortion". That defnition is left to the
discretion of the inspector.
EngAddict - you are incorrect.
Quote (UG-22(j))

The loadings to be considered in designing a vessel shall include those from:


(j) test pressure and coincident static head acting during the test (see UG-99).

and UG-99 does not have a limit as discussed above.


jtseng123 (Mechanical)

13 Sep 12
14:23

Talking about hydrotest, now come to my real case just happening:


I have a 40 feet long vessel, 56" diameter, one half is CS at 60 psig/650F refractory lined, the
other half is SS at 60 psig/ 1450F.
Per code, test pressure for CS section =100 psig, but can be tested up tp 500 psig so the
primary membrane stress is 90% of the yield.
Test pressure for SS section = 1,000 psig.
How am I going to run the hydrotest ? Test separately at their own test pressures and then
weld together ? or weld together and then test at the highest pressure it can take (500 psig) ?
or test entire vessel at 100 psig ?
TGS4 (Mechanical)

13 Sep 12
14:24

I should have qualifed my above post that my quotations are from the 2011 Edition of VIII-1.
However, looking back to previous Editions/Addenda, back in 2007, the wording was identical
except that the prohibition on 1.3xLSRxS_bolt(@ test temperature) > 0.9xSy was not there
(that specifc item was added in 2010). Going back to the 1998 Edition (when the ratio was
changed from 1.5 to 1.3, there was still no limit of 0.9xSy. The earliest edition that I have onhand (1980) does not have the limit.
(On an historical note, even the 1962 Edition of Section VIII, had no mention of 0.9xSy)
My guess is that the use of 0.9xSy has crept in from the old (pre-2007) versions of ASME
Section VIII, Division 2. There, article AD-151.1 stated

Quote (ASME Section VIII, Division 2 (2004 Edition), Article AD-151.1 For Hydrostatically Tested
Vessels)

The hydrostatic test pressure of a completed vessel shall not exceed that value which results in the
following stress intensity limits:

Confidential 2012 Aker Solutions

Page 3 of 11

(a) a calculated primary membrane stress intensity P_m of 90% of the tabulated yield strength S_y at
test temperature as given in the applicable table of Subpart 1 of Section II, Part D...

Note that this limit is different in the current Edition of Division 2 - see article 8.2.1 and
4.1.6.2(a)(1).
But, to conclude, there is no such limit for vessels constructed to ASME Section VIII, Division
1.
EngAddict (Mechanical)

14 Sep 12
4:02

@TGS4, my comments were regarding requirement to consider the design for the hydrotest
condition, which it does. I was not suggesting it gave this allowable limit, only that it must be
considered, and where the code does not specifcally cover details of the design it does not
mean it can be excluded. In such absence you can refer to other codes (or standard industry
practice) for this guidance provided the method is as safe as those provided by the rules of
division 1. Yielding around structural discontinuities is permitted and some distortion is
acceptable since this is a serviceability limit state not a strength limit state. My concern would
be regarding exceeding yield in general primary membrane away from discontinuities.
@jtseng123, it is based on the 'lowest' stress ratio of all materials of construction. Must be
tested as a complete vessel, after welding.
prex (Structural)

14 Sep 12
4:51

EngAddict, UG-22(j) is simply reminding the designer that in some cases (specifcally tall
vessels where the process fluid is not liquid) the hydrotest with the water head added may
represent a limiting design condition.
I agree with TGS4 that there is no mandatory per code stress limit on the hydrotest stress,
and also even that there is no obligation to submit calculations for the hydrotest condition
except where that condition may be limiting, as per UG-22(j). The only further example that
comes to my mind of a hydrotest limiting condition is when a tall vessel is tested in horizontal
position, the test pressure is calculated including the liquid (operating) head, but the upper
portions of the vessel, that do not see the operating liquid head, might go overstressed.
It is however evident that going beyond yield (with a reasonable margin) on a general
membrane stress would be a source of troubles, so any designer would check this in case of
doubt, but I can't see how this could happen apart from the special situations mentioned
above.
jtseng123, your test pressure seems to be 100 psig, but you are free to propose to your
client to go up to 500 psig for the sake of safety. As pointed out by TGS4, there is no
mandatory upper limit on test pressure per VIII 1.
prex
http://www.xcalcs.com : Online engineering calculations
http://www.megamag.it : Magnetic brakes and launchers for fun rides
http://www.levitans.com : Air bearing pads

EngAddict (Mechanical)

14 Sep 12
6:45

prex, I don't think your comments contradict mine. Technically UG-22 j) applies to all vessels
but will only affect some limited cases, for most vessels the hydrotest pressure will always be
acceptable. I agree with your comments, I would not request calculations unless my checks
found that the test pressure is limited by design. The designer must consider all loading
conditions that affect the vessel, whether they then rule them out as a limiting factor or not is
beside the point.

Confidential 2012 Aker Solutions

Page 4 of 11

TGS4 (Mechanical)

14 Sep 12
12:18

EngAddict - I have no disagreement with you regarding that all loads during the hydrostatic
test must be considered. The example that you provided is exactly the situation where this
paragraph is important. However - the question that you answered was in regards to this
supposed (turns out to be fctitious) 90% of yield limit. As I stated before, there is no specifed
upper limit on hydrostatic test pressure in Division 1.
jtseng123 - As EngAddict pointed out, the St/S ratio is based on the LOWEST ratio of the
materials of construction. Assuming SA-516-70 for your CS (S=20ksi@70F and
S=18.8ksi@650F) and SA-312 TP304H for your SS (S=20ksi@70F and S=1.8ksi@1450F),
your LSR would be 1.064. So, your hydrostatic test pressure would be 60psi x 1.3 x 1.064 =
83 psi.

All - I am going to keep repeating this until I blue int he face - there is NO "90% of yield" limit
in ASME Section VIII, Division 1. Stop using it. Stop referring to it. Stop.
Furthermore, in ASME Section VIII, Division 2, there is such a limit, as I discussed above
(Article 4.1.6.2(a)(1)). That limit is now 95% of room temperature yield. It was 90% but is now
95%. So, please stop using 90% of yield. It doesn't exist in any current Code and hasn't for
the past 5 years.
Please erase from your memory any reference to 90% of yield for hydrotest.

vpl (Nuclear)

14 Sep 12
12:54

TGS4 I gave you a star -- but I should point out that your little soapbox man is actually quite
red in the face rather than blue. You make a very good point that the Code has been updated
and people need to be using the latest version.
Patricia Lougheed
******
Please see FAQ731-376: Eng-Tips.com Forum Policies: Eng-Tips.com Forum Policies for
tips on how to make the best use of the Eng-Tips Forums.
jtseng123 (Mechanical)

14 Sep 12
13:55

since my vessel has two signifcant design temperatures, if only test at 100 psig, or 500 psig
based on CS section, that will not stress the SS section enough to see any leak so to simulate
the design condition at 1450F. Does code ever address how to hydrotest for two signifcant
design conditions ?
TGS4 (Mechanical)

14 Sep 12
16:33

Quote (jtseng123)

...that will not stress the SS section enough to see any leak...

Just curious - why do you think that is the purpose of a hydrostatic test? Please refer to UG-

Confidential 2012 Aker Solutions

Page 5 of 11

99(g). The hydrostatic test is neither a proof test nor a leak test - although a leak check is
required at a reduced pressure (UG-99(g)). The Code is silent on the rationale for performing a
hydrostatic test. As an aside - that is not for lack of trying, BTW. There is not general
agreement within the various Code Committees themselves about the purpose of a
hydrostatic test.
jtseng123 - this situation is rather common and is one that is fully anticipated by the Code.
Every other vessel that has followed these rules has turned out OK - why would you think that
yours would be different?
vpl - thanks. There is no emoticon with a blue face
jtseng123 (Mechanical)

14 Sep 12
22:25

TGS4 and others: the term 'leak test" though not technically correct, but is commonly used
across the entire industry, no need to picky on it. It will be still used for the next hundreds of
years.
Can anyone explain why you have to time the stress ratio ? My understanding from
mathematic point of view is, the stress ratio is to simulate the design condition, especially the
design tempertaure that is impractical to heat it at shop to check if the vessel can sustain or
not. In other words, it is to "fully stress" the vessel at ambient temperature. 1.3 factor, or 30%
more, is just to stretch the vessel a little bit more, and if it won't fail, your vessel will be good
at design condition plus additional 30% safety factor. Due to this 30% extra, code allow you to
back down your test pressure by using the lowest stress ratio, or if the primanry stress exceed
95%, 90%, yield, or whatever other limits that have been talked about so much in this post,
you can reduce the test pressure without causing hiccup.
simple math for hoop stress divided by allowable: P1*R/t/S1 = P2*R/t/S2 = 1(plate is fully
stressed)
where P1=test pressure, S1=allowable stress at test temp
P2, design pressure, S2 = allowable stress at design temp
t , wall thickness, R: vessel radius.
So, P1=P2*S1/S2, now you see how the stress ratio appears there. Code just further adds 1.3
factor to it.
If stress ratio variation is within 30% for the entire vessel, it will be fne according to the math.
In my case, the stress ratio variation is from 1 to 11. If I only test at 100 psig and in case
there is defect in SS section, lack of fusion, internal crack, that is not detect, but can
withstand 100 psig test pressure, does anyone think it won't burst when heating up to 1450
at 60 psig where the allowable stress is so low ?
TGS4 (Mechanical)

16 Sep 12
1:35

Burst - no. The failure mode at those temperatures is creep. The allowable stresses are based
on lower-bound material properties - most materials perform much better.
Your exact situation has been discussed at the Code Committees tables. The writers of the
Code are well aware of this possibility. They are OK with it. As I said - the purpose of the
hydrostatic test is NOT necessarily to stress that material to 30% beyond the value that it will
see at design. I get your math, though.
For your particular situation, remember that Division 1 does not place an upper limit on the
hydrostatic test pressure. If you'd be happier with 500 psig, then there is nothing stopping
you. And don't worry about exceeding yield in primary membrane - carbon steel has
signifcant strain hardening. If you ran an elastic-plastic analysis of your hydrostatic test
condition (at the pressure you want to test the SS part to) you'd probably fnd that it won't

Confidential 2012 Aker Solutions

Page 6 of 11

actually fail. If your simulation didn't show deflections exceeding about 1/2", I would likely say
that that passes the "visible permanent distortion" test...
jtseng123 (Mechanical)

16 Sep 12
14:10

TGS4, thnaks. However, not so clear to me.


Let's disregard the 1.3 factor which to me is beef added by code, and also forget how good
the modern materail is.(Note that base metal can be excellent, but welding is still controlled
by human, so the hydrotest is to test more of the workmanship than base metal. though base
metal does burst, but not so often. I use "burst" is from the failure appearance I have seen so
many of them. No need to picky the word I use. Many of our overseas projects, the frst piority
is to get a good welder to minimize weld rejection rate),
just look at the simple hydrotest equation P1=P2*S1/S2. To me, it is to "simulate" the design
condition from the math to make sure the vessel can "sustain" even there is defect in welding
and base metal. (weld defects may exist because code allows spot or even no NDE). So as
long as there is no leak, regardless how bad the weld and base metal are, the vessel is good
to go for operation, becasue it pass the "simulation" test. (same as the fnancial meltdown
that banks must pass simulated "stress test" to continue operating). That is my
understanding of the purpose of hydrotest, straightforward with the technical basis by math.
UG 99(g) saying inspection pressure shall not be less than the test pressure divided by 1.3,
that exactly hits the math, and 1.3 is additional safety factor code added.
You have not clearly stated the purpose of hydrotest, would like to know your interpretation if
other than mine.
In my case, it is not I am happier with 500 psig or not. I am the person to tell vendor what to
do in this situation, and every decision I make, I must have technical basis. With 500 psig or
higher that is way above the required for CS section, if something goes wrong, I will be
responsible becasue "I tell vendor to do what is not desgined for". If tested at 100 psig, that
means SS section is dramatically not pressure enough to "simulate" the design condition,
then I have no techncial basis to state it will not fail when put to service.
Any knd of analysis, elastic-plastic, deformation, FEA, etc, you name it, can not substitute the
real world hydrotest. That is my years experience in this feld, and that is why code ask to do
hydrotest after you have done your calcualtion and even with 100 % NDE on every weld and
100% UT on all base metal.
If my interpretation for the purpose of the hydrotest is correct, code committees shall take a
second thought, maybe add limitation on the variation of stress ratios. And if variation is too
large, just like my case, recommend what shall be done. I tends to ask vendor to test SS
section separately at 1000 psig. Then weld CS and SS together and test at 100 psig entirely.
The disimilar closing girth seam will be covered by refractory and set to 650F, so no need to
worry.

TGS4 (Mechanical)

17 Sep 12
12:09

jtseng123 - I think that the hydrostatic test has many purposes - one of them is what you
describe above. I completely agree that what you describe above is one of the purposes.
Quote (jtseng123)

If my interpretation for the purpose of the hydrotest is correct, code committees shall take a second

Confidential 2012 Aker Solutions

Page 7 of 11

thought, maybe add limitation on the variation of stress ratios. And if variation is too large, just like my
case, recommend what shall be done. I tends to ask vendor to test SS section separately at 1000 psig.
Then weld CS and SS together and test at 100 psig entirely. The disimilar closing girth seam will be
covered by refractory and set to 650F, so no need to worry.

I think that you may have a valid point about limiting the variation in St/S. Please submit an
inquiry to the Code Committees, as outlined in "Submittal of Technical Inquiries To The Boiler
and Pressure Vessel Committee - Mandatory" in the front matter of the Code.
jtseng123 (Mechanical)

24 Sep 12
22:53

My USA vendor intimidated me today if I insist hydrotesting C.S section and SS section
separately, then weld together, it will have cost impact plus delay.
I told vendor, prove to me that testing the SS section at 100 psig and then put to service, if
not causing failure, I will buy it.
I futher told vendor why not just put miminum welding deposite, and have all kinds of
crack/voids in the base metal that nrmally does not receive NDE, even have crack/voids in the
welds that NDE is not required, and I can guarantee 100 psig test pressure can be easily
passed. And if he dares to put into service for 1450F/60 psig becasue he says it passes "100
psig" test pressure, I can guarantee him it's going to burst the frst minute in service.
doct9960 (Mechanical)

25 Sep 12
0:46

jtseng,
If what you require is not in the contractual agreement, then you have to pay extra and
expect a late delivery. That is not intimidation, that's business. Nothing is free.
If you did not specify any special hydrotest requirements in your PO, the vendor will use UG99(b) for the hydrotest pressure.
What is the MAWP that will be stamped on the vessel nameplate? And what are the stress
ratios for the CS section and SS section?
jtseng123 (Mechanical)

25 Sep 12
17:52

doct, the stress ratio is 1 for CS and 11 for SS. Actually, this is a long 40' duct, 56" diameter
per B31.3. But for hydroetst, not much difference between B31.3 and Div.1. B31.3 has 1.5
factor in liu of 1.3 in Div. 1.
If you carefully study how the test pressure formula comes from and its purpose, you will see
my point. I believe TGS4 has seen my point. So, basically, I disagree with you. You can not
test a major piece of equipment at 100 psig, which supposedly to be tested at 1000 psig in
order to see any meanfully defect. Is hydrotest vendor's responsibility ? yes it is. Shall vendor
flag this special problem ? maybe they should when they bid the job and if they understand
the formula. The truth is, hardly any vendor or anyone facing this kind of problem in their
lifetime. Code is not all-powerful to cover all situation.
doct9960 (Mechanical)

25 Sep 12
23:23

jtseng,
You mislead a lot of us in this forum into thinking that your "vessel" is designed to ASME VIII1, when in fact it is just a piping component. When TGS4 mentioned that "you may have a
valid point about limiting the variation in St/S", he was referring to the ASME VIII-1.
There are a lot of differences in the hydrotest requirements of ASME VIII-1 and ASME B31.3.

Confidential 2012 Aker Solutions

Page 8 of 11

ASME B31.3 has a limit on the stress ratio; it should not exceed 6.5. There are also
recommendations of reducing the hydrotest pressure if stresses exceed the material yield
sstrength or 1.5x the component rating. ASME B31.3 is also specifc on what hydrotest
pressure to use if the piping system has more than one material or more than one design
temperature. I suggest you get the latest copy of ASME B31.3 and carefully read paragraph
345.4.
So with a design pressure of 60 psig and assuming you got your stress ratios right (based on
Table A-1 or based on component rating), your minimum hydrotest pressure would be 60 x 1.5
x 6.5 = 585 psig.
If you did not specify a 1000 psig hydrotest pressure in your P.O. and if you did not specify
separate hydros for the CS and SS section, the vendor has every right to charge you extra
and delay delivery.

jtseng123 (Mechanical)

26 Sep 12
12:15

Doct, thanks for pointing out 6.5 limit but I do not know what is the basis for that. The rest
you talking about, yield and component limitation, etc, that I fully understand.
There are issues on this that I concern:
1. The integrity of the SS section that is not tested to its own presuure to simulate design
condition. Even I discount the 1.5 muiliplier, the SS still must see 60x11= 660 psig test
pressure for a meaningful hydrotest to guarantee it will be no problem in service.
2. Unneccessary testing the CS section at 585 psig for which it is not designed for, induces
other risk. There are many manways and nozzles on the duct, so all reinforcement have to be
rechecked, redesigned, refabricated. That for sure will cause cost and delay. I do not think it is
wise to test at that high pressure for CS section.
So, comprimizing everything and then testing the system at low pressure becasue people
keep referring "code" has not much meaning at all in my opinion becasue it defeats the
purpose of hydrotest for SS section.
If you really have the project I have, are you willing to allow vendor to test at low pressure for
SS section and you can endorse it " hey, SS section won't be a problem". And what will be
your technical basis on that ? Becasue code says so ? Well, it's better to fnd out what's
behind the 6.5 ratio limit and what kind of materal is involved. Is it based on the same
dedesign temperature ? or similar material group ? Has code included my extreme case ?

TGS4 (Mechanical)

26 Sep 12
12:39

jtseng123 - you have a problem with the Code itself.


Quote (tgs4)

Please submit an inquiry to the Code Committees, as outlined in "Submittal of Technical Inquiries To The
Boiler and Pressure Vessel Committee - Mandatory" in the front matter of the Code.

Since your case is really about B31.3, then follow the procedure in Appendix Z - Preparation of
Technical Inquiries.
We can argue here all we want, but your beef is really with the Codes. Take it up with them.
When you have submitted your inquiry (inquiries), please inform us here - many are Code

Confidential 2012 Aker Solutions

Page 9 of 11

Committee members, or work with Code Committee members. We may take an active role in
shepherding your inquiry through.
The Code Committees have the background and history (and access to the history) of these
requirements. Maybe there was a technical reason. Maybe there is test-data to back up what's
there. Maybe it's completely arbitrary with no technical backup. Maybe it's a compromise
position due to the consensus nature of the Committees. Maybe...
jamesl (Mechanical)

27 Sep 12
9:50

It is common to run into a situation jtseng123 described when you use dissimilar materials on
one vessel. For example, before the Code made revision in UG-99 several years ago, it was
common to have bolting material for custom designed body flange limiting hydrotest
pressure.
Hydrotest can be a test of structural integrity, leakage check and serves as mechanical stress
relieve. I agree we should stress the vessel as more as possible, to 90%, 95% or even 100%
yield so as more defects can be exposed as possible, if there is any. However we should
realize it is almost impossible to stress test every weld seam to its maximum in a vessel
during hydrotest. Think about those weld seams in a large nozzle neck or weld seams
between flange and nozzle neck. The stresses there are not going to be very high even during
a hydrotest.
Jtseng stated:
If I only test at 100 psig and in case there is defect in SS section, lack of fusion, internal
crack, that is not detect, but can withstand 100 psig test pressure, does anyone think it won't
burst when heating up to 1450 at 60 psig where the allowable stress is so low ?
If the stainless steel part of the vessel is stressed to 2600psi (I just use a random number
here) at hydrotest without problem, is it unacceptably dangerous to stress it to 2000 psi @
1450 F. I doubt it. Why would it burst? The safety factor built in the Code is still there in
addition to the 30% higher stress during hydrotest. Besides, as pointed out by TGS4,
structural failure at hydrotest temperature and at 1450 deg F would be a totally different
mechanism any way.
If the carbon steel part of the vessel can be tested at 500 psi, why not just test the vessel @
500 psi and be done with it?
jtseng123 (Mechanical)

27 Sep 12
23:34

Very good. Everyone has his opinion addressed. Here is the reality check: After disscussed
with many heavy weight engineers, inside and outside the company, all in agreement that SS
section shall be tested separately at its own high test pressure (585 psig), then welded to CS
and do the second test at the CS low test pressure (112 psig). No one wants to test SS at
pressure lower than 585 psig because there is no technical basis.
The other option is: test this SS/CS hybrid together at 585 psig to satify SS section, and beef
up reinforcement on CS section for this high test pressure that it was not originally designed
for.
I select the latter one and advise vendor to proceed. Be done with it.
I have the right to decide which code to be used. Fortunately, I select B31.3, not Div. 1. If I
select Div. 1, I will be in much trouble to fght with the low stress ratio with 100 test psig on
SS that wholy defeats the purpose of hydrotest.
This SS/CS hybrid is out of the code. And good engineering judgemnet must be called for.

Confidential 2012 Aker Solutions

Page 10 of 11

If everyone still has interest, here is a simple question:


A pure 304H SS equipment at 1450F/60 psig, can be either B31.3 or Div. 1,
Per B31.3, test pressure is 585 psig because code says stress ratio to be 6.5 maximum.
Per Div. 1, test pressure is 860 psig because there is no stress ratio limit except yield.
Why so much difference in test pressure ?

Confidential 2012 Aker Solutions

Page 11 of 11

You might also like