EN13445 Fixed Tubesheet Loading Conditions 1

[FPPE]

Dears,

I would like to understand which temperatures to consider according to the following paragraph of EN 13445 for the calculation of fixed tubesheets heat exchangers:

The code mentions normal operating conditions, the start-up conditions, the shut-down conditions, the upset and the pressure test conditions. Our software automatically defines the following temperatures (as wall temperatures) and the following loading cases (based on the 4 temperatures filled in input):

For each of these conditions it performs the calculation correctly by differentiating for the 7 loading cases.

My first doubt concerns the upset conditions: what temperatures should be considered? Are these conditions mandatory per EN 13445?

Second doubt: according to what I have found in this article

https://www.sant-ambrogio.it/en/wha...alculation-of-shell-and-tube-heat-exchangers/

, one should always specify (based on the output of the thermal calculation) the actual wall temperatures defined on the basis of the properties of the fluids and how they actually transmit heat. From this I understand the simplicity (not always having the full thermal calculation available) of averaging the 4 temperatures, which is certainly not correct. How do the software you use take into account the temperatures in the table in our software?

Third doubt: we have to design the exchanger having the following temperatures

SHELL SIDE IN/OUT 40/219.4
TUBE SIDE IN/OUT 257/105

Design pressure = 30 barg + FV each side
Design temperature = 280 °C each side

I have designed more than 50 exchangers with fixed tubesheets, but now I cannot find a combination of thicknesses that satisfies all the equations in the code, so all the doubts about these loading cases have arisen.

Can you help me?

Thanks in advance

 

[SnTMan]

FPPE, I am not familiar with EN 13445, but based on your excerpt it looks similar to ASME Sec VIII, Div 1 Part UHX rules.

Your doubt 1), any upset conditions would, in my experience, need to be supplied by the client. These would be conditions of abnormal flow or temperature excursions. It is normally not possible for the designer to accurately identify such conditions from the heat exchanger data sheet. Don't know if it is mandatory in EN 13445. Pretty much the same language appears in UHX, but consideration of upset conditions was not mandatory thru 2017 Edition.

I would generally agree with the approach used by your software as to the temperatures used in the start up, operating and test cases, although start up cases might sometimes be evaluated at Minimum Design Metal Temperature (MDMT) rather than ambient. I am less sure about the shut-down cases, it seems to assume both sides are shut down at the same time. Could be true.

Your doubt 2), it is usually preferred for the operating cases to use temperatures from the rating data, what you call wall temps, which I assume is what UHX calls mean metal temperatures. However, HX's must sometimes be designed in absence of detailed rating data, in which case the temperatures called out in Notes A & B seem reasonable to me. Software I was familiar with required entry of the MMT's as input. If no detailed rating data was available average of inlet & outlet temps was normally used. Does your software allow wall temps / MMT's as input?

Your doubt 3), which case(s) are you failing on, and in which ways? Tubeside start up can often be a most severe case as the tubes are expanding while the shell is not.

The problem with sloppy work is that the supply FAR EXCEEDS the demand

 

[Christine74]

HTRI provides estimated Mean Metal Temperatures (MMTs) for each tube pass. TEMA also has guidelines for calculating tube/shell MMTs.


-Christine

 

[FPPE]

SnTMan thank you for the detailed answer.
Doubt 2) Yes, the software allows the input of MMT.
Doubt 3) Tubeside start-up is the case that gives me problems, since Tin tubeside equals 257 °C.
Connecting to Doubt (2), I had to input MMT temperatures during start-up.

Typically, in my experience, the cold side is the one that is started up first (in this case the shell side), so I assumed the MMT shell side equal to the average of the shell side temperatures and tube side the average of T inlet and outlet (tube side).
This is because, considering the nature of the fluids flowing and given their heat transfer properties, the tube side fluid (mostly gaseous) should have a temperature closer to the one flowing where the state is liquid (shell side, see attached datasheet), so the actual metal temperature should be closer to the average temperature between inlet and outlet (to be conservative, in reality should be closer to shell side temperatures), since the shell side is running at tubeside start-up.

Christine74, thank you very much. Probably the best thing to do in this case to be sure is to run a thermal analysis with HTRI during start-up (with shell side already started).

 

[SnTMan]

FPPE, I think a start-up condition is one with flow at inlet conditions one one side with no flow on the other. This corresponds with the table from your software. Given that flow is present on one side only, little or no heat transfer takes place.

Your statement "... I assumed the MMT shell side equal to the average of the shell side temperatures and tube side the average of T inlet and outlet (tube side)" can only be true if flow is present on both sides of the exchanger.

Shell start-up is rarely a problem even with the tubeside at an MDMT below ambient. This is because both the shell and the tubes are essentially at the same temperature, since both are in intimate contact with the shell fluid, therefore very little differential thermal expansion.

Tubeside start-up is the opposite case. Since flow occurs only in the tubes, the fluid temperature is only slightly reduced from inlet, if at all. Tubes are expanding, the shell is not. Differential thermal expansion is important. Tube allowable compressive stress and / or tube-tubesheet allowable joint load may be exceeded.



The problem with sloppy work is that the supply FAR EXCEEDS the demand

 

[FPPE]

SnTMan, you are right, thank you. I solved this by inserting an expansion joint.
Kind regards

 

[SnTMan]

FPPE, expansion joint, necessary evil. Yuck

Best of luck.

The problem with sloppy work is that the supply FAR EXCEEDS the demand

 

[FPPE]

The dark beast of pressure vessel engineers, thank you!

 

[r6155]

Try to avoid the expansion joint.
This difficult manufacturing, inspection, pressure test...
Be careful with the saddle supports, thermal insulation, transport, ......... etc.

Regards