Shell and Tube heat exchanger design temperature 2

[Dragan12]

thread124-390770

Hi all,

As far as I have seen on this forum there are different opinions regarding design temperature for shell and tube HEx. Some say that it is ok to have same design temperature for shell and tube side, some say that it should be different. Main concern are the tubes whose op. temp. and consequently design temp. can be larger than the design temperature of other tube side elements, e.g. channels. This is valid under assumption that hot fluid is on the shell side.

I found in ASME VIII, Div. 1 UG-20 DESIGN TEMPERATURE following statement: "The design of zones with different metal temperatures may be based on their determined temperatures."

This means that I can specify one design temperature for tubes and other for other tube side elements of HEx. Is this correct or am I seeing something wrong?

Thanks in advance.

 

[georgeverghese]

Upper design temperatures should be the same for both tube and shellsides. And the same should apply for the setting of the lower design temp also.
This principle will be the inevitable outcome on consideration of the operating range of flows down to min turndown, flow transients resulting in control loop upsets, and at least one operating and coincident trip loop failure scenarios.

 

[Dragan12]

Than one should also consider updating connecting lines design temperature in accordance with design temperature of the HEx?

There is no sense to have e.g. design temperature for tube side 200C and for IN/OUT lines 100C.

 

[jte]

For ASME VIII, the design temperatures can vary from component to component as you note above. For example your tube design temperature can be different than the channel's design temperature. Now, a practical calculation or measurement will be difficult to come by, and you'll need one to justify the different (fine-tuned) temperatures. Don't forget to evaluate all scenarios, including when one side is shut in while the other side remains in service. Not merely "normal" operating conditions. Sometimes valves get closed which weren't meant to be closed, and this alone should not cause the failure of the exchanger. More often we'll see varying design temperatures in distillation columns where calculations and experience tell us that the temperature at the top will be substantially different than the temperature at the bottom during all operating scenarios.

 

[Dragan12]

Its clear that due to some operator error or trip failure "cold side" can see higher than normal operating temperatures, but does that mean that HEx "cold side" outlet line should be designed for this scenario?

I apologize if I went a little bit off-topic, but I think that this is also important during plant design.

 

[EmmanuelTop]

I would say - yes, on your last question. Normally the temperature specification break for piping will not be on the exchanger outlet flange, but on the farthest isolation valve in the discharge piping. This is what the common logic would dictate.

API RP 14J provides some guidance on pressure specification breaks. Analogy can be taken for temperature spec breaks.

Dejan IVANOVIC
Process Engineer, MSChE

 

[georgeverghese]

Can you describe your HX configuration, operation, valving on both hot and cold sides and the other equipment it is connected to, together with a control scheme ? - this will help to determine if what you say is adequate from a process safety aspect.

 

[Dragan12]

Here is one example in the attachment.

 

[don1980]

The design T for the shell side is based on the temperature range that the shell side will experience, plus a reasonable addition for safety factor. The same is true for the tubeside. Those are independent decisions. That is, the shell side design T isn't a factor when determining the tubeside design T, and vice versa. These design T values are seldom the same number, nor do they need to be the same. The mechanical designer will evaluate the stresses caused by expansion/contraction due to the specified design T values. If the stresses are too high, and the design T values can't be changed, then one or more expansion joints are needed. That's a core part of the mechanical designer's overall task.

 

[georgeverghese]

Good of you to post this sketch - that makes things clearer.

This is a typical case where design temp for the shellside must be made equal to that for tubeside to account for the case when hot oil flows through the HX with little or no crude flow, which will happen when the hot oil SDV is nominally closed by action of the TSHH or by PSD or manually stopped, but is still leaking.

The TSHH setpoint can be used to derive the design temp for the crude side piping outside of the isolation block valves for the crude side.

A complication arises when in this case, hot crude inventory flows out of this HX at 190degC to the stage 2 sep during a restart after shutdown - the tubeside crude hot inventory flowing out should not raise the metal piping mean metal temp by much - take this into account when setting the design temp of the crude piping after the HX crude side downstream block valve. Heat transfer coeff from the flowing hot crude to the piping will also be low during this restart transient. Also check if the crude side LCV trim can handle this short term temp spike.

 

[EmmanuelTop]

In addition to the above:

- Failure of the hot oil temperature control loop (TIC) with TCV in fully open position (full flow of hot oil) results in crude oil flowing to equipment downstream at a higher-than-normal (uncontrolled) temperature. This can last for undefined period of time (no visible safeguards), hence it will define the design temperature for piping and instrumentation downstream of the heat exchanger.

SDV can also be moved closer to the exchanger, upstream of the level control valve and isolation valves.

Dejan IVANOVIC
Process Engineer, MSChE

 

[Dragan12]

@georgeverghese

"This is a typical case where design temp for the shellside must be made equal to that for tubeside to account for the case when hot oil flows through the HX with little or no crude flow, which will happen when the hot oil SDV is nominally closed by action of the TSHH or by PSD or manually stopped, but is still leaking. "

When there is no crude oil flow PSD is beeing activated and Hot Oil Heater Unit burner and hot oil circulating pumps are stopped, therefore there is no flow in the hot oil system.

This other case with restart after shutdown is possible, but it can be difficult to predict (calculate) temperature spike downstream.

@EmmanuelTop

"Failure of the hot oil temperature control loop (TIC) with TCV in fully open position (full flow of hot oil) results in crude oil flowing to equipment downstream at a higher-than-normal (uncontrolled) temperature. This can last for undefined period of time (no visible safeguards), hence it will define the design temperature for piping and instrumentation downstream of the heat exchanger."

If TIC fails there is a shut down valve on hot oil inlet connected to TIT on crude outlet with HH switch. This HH switch should be something above TIC high value and should serve as a safeguard system if TIC fails.


Thank you all for replies and time invested, you have shed a new light (at least for me) on this topic.

 

[georgeverghese]

Dragan12, Agreed that the only premise for unequal design temp, from an equipment failure point of view, would be for the hot oil pumps to be stopped for all cases when this HX is shutdown.

But you also have to account for the case of misoperation where the operator is flowing hot oil through the HX with no crude flow during startup. In most companies, this risk is significant.

Also, it is common practice to keep the hot oil pumps running for some time after a heater shutdown to cool down the radiant bank tubes in the furnace and dissipate the residual heat in the refractory bricks, and also to stop the trapped hot oil in the heating coils from coking up. So check again if this trip of the HO pumps is advisable. Stopping the HO pumps is only required for furnace exhaust stack TSHH, which may be an indication of a leak in the heating coils.

If the HO pumps are still kept running to protect the furnace coils and to prevent hot oil thermal degradation, leakage through this SDV is still possible since the hot oil will still be at 190degC for some time even though the furnace burners are stopped.

 

[Dragan12]

Its true that in order to prevent thermal degradation of hot oil there has to be some min flow through heating coil in fired heater. Maybe this TSHH should be removed in order to, as you say, enable this min flow.

There is also a question of SDVs closing sequence during ESD and PSD. If all SDVs on hot oil distribution lines close in the same time when fired heater burner is turned off there is no min flow for fired heater. Hot oil pumps would keep some flow through automatic recirculation valves but this action would keep the flow only for pumps, not for coil inside fired heater. Therefore some time gap between burner shut down and SDV closing should exist.

 

[georgeverghese]

Or maybe re arrange the SDVs' and relocate the min flow recycle line to allow min flow to occur through the fired heater on PSD. The TSHH is the only protection you have for upper design temp of the stage 2 sep and its feedline.

 

[Dragan12]

But what happens when I redirect min flow of oil heated to 190C back through fired heater. For me this is a question for Vendor.

 

[georgeverghese]

If ht oil is already at 190degC, then burner should be at min stop firing load which should ideally be a very low firing load. Most fired heaters cannot deal with a very low firing load, hence a trim cooler is installed in this HO circulation loop to dissipate excess heat produced by the burners when there is no process demand. This prevents having to stop the burners altogether. Trim cooler design load should dissipate all the heat produced by the burners when at min stop firing.

 

[EmmanuelTop]

Design with an air cooler on the recirculation (bypass) line is quite common - we have operated such units in three different gas plants worldwide. I need to say here that the vendor does not necessarily know what system and what requirements you have, so the air cooler will probably not be in the base case design option (see the links I posted below) - it comes as a conclusion from operability and safety assessments - such it is an example in this thread - and it falls on you (= your company) to communicate this requirement to the vendor.

See also:

http://www.process-heating.com/articles/84799-tips-hot-oil-systems

http://igs.nigc.ir/STANDS/IPS/e-pr-410.PDF

https://www.icheme.org/~/media/Docu...vention/Hazards Archive/XXI/XXI-Paper-025.pdf

Dejan IVANOVIC
Process Engineer, MSChE

 

[georgeverghese]

Typically, the trim cooler is an air cooled unit (avoids scaling issues with water cooled units) and is set to dissipate approx 25-30% of the rated burner capacity at the prevalent normal operating burner fuel gas supply pressure. The fans are auto switched on to operate at a given temp gap, and auto stop when the op temp is below this gap.