Jump to content



Featured Articles

Check out the latest featured articles.

File Library

Check out the latest downloads available in the File Library.

New Article

Product Viscosity vs. Shear

Featured File

Vertical Tank Selection

New Blog Entry

Low Flow in Pipes- posted in Ankur's blog

- - - - -

Designing Flare Systems For Low Temperature


This topic has been archived. This means that you cannot reply to this topic.
25 replies to this topic
Share this topic:
| More

#1 Butterfly

Butterfly

    Gold Member

  • Members
  • 73 posts

Posted 20 April 2012 - 10:02 AM

Dear all

If I want to design the flare system (headers, etc) I have to take care of the low temperature that will be reached during depresurization, to use the right materials. Wind velocity will have some influence in this: if it is high, it will help to heat the flare header, so I guess that the conservative design should take into account the lowest wind velocity (for the low temperature case).

Is this right? If so, is there any standard which indicates this, that the design shall take into account the lowest wind velocity?

Thanks

#2 Nick Howard

Nick Howard

    Brand New Member

  • Members
  • 2 posts

Posted 20 April 2012 - 11:20 AM

Butterfly,
Are you perhaps confusing pressure let down and dispersion here? Yes, if you were to let down your pressure into a flare header then you would get an accompanying drop in temperature. In addition, you could include solar gain in your calculations.
When you flare, you would need to look at the radiation isopleths for their effect on people and equipment. However, if you were to cold vent (or fail to ignite the flare) then you would get a nice flammable cloud to disperse.
Hope this helps, but I'm not sure it will!
Regards,
Nick

#3 kkala

kkala

    Gold Member

  • Banned
  • PipPipPipPipPip
  • 1,939 posts

Posted 20 April 2012 - 04:42 PM

To my understanding of the case, any heating effect of wind should be neglected, resulting in a conservative design. This is what we apply here, other designs comply with it.
For instance, suppose that you depressurize propane. Propane boiling point at atmospheric pressure is -42 oC, so gas of -42 oC enters the header. Design temperature of header should be -42 oC.
I assume that -42 oC can be developed for some time inside the header, irrespectively of the external wind.

Edited by kkala, 20 April 2012 - 04:48 PM.


#4 ankur2061

ankur2061

    Gold Member

  • Forum Moderator
  • 2,484 posts

Posted 20 April 2012 - 11:55 PM

Butterfly,

The design of the low temperature should take into account the following:

1. Provide controlled depressurization in order to prevent the flare header temperature to fall below -29°C. Below -29°C low temperature carbon steel (LTCS) would be required to prevent brittle fracture of ordinary CS.

2. Flare streams of cold dry gas and wet moist gas shall be segregated thereby avoiding the possibility of freezing and hydrate formation. If a cold stream is vented to the flare header, the header will be cold. If subsequently a warm moist gas is released, hydrates or ice could be formed which could block the header.

3. A cold stream could be routed through a separate header to the common Knock-Out Drum (KOD) including a common flare system downstream of the KOD.

The idea of a flare system design would be that while avoiding separate flare systems for one plant or unit the aspects of operability and safety of the flare system are fully evaluated before proceeeding with the design. This would involve a careful study of the effects of mixing of incompatible streams in the flare system.

Hope this helps.

Regards,
Ankur.

#5 kkala

kkala

    Gold Member

  • Banned
  • PipPipPipPipPip
  • 1,939 posts

Posted 21 April 2012 - 09:43 AM

One clarification on controlled depressurization to prevent flare header temperature from falling below -29 oC would be useful.
Suppose that liquid propane is depressurized through a system of block valve and downstream orifice. No control valve should be placed between orifice and header. Pressure upstream of orifice (same temperature as downstream of orifice) continuously falls to 6 Barg, reducing temperature to 13 oC. Then depressurizing continues, until pressure in the vessel gets same as pressure in header, close to 0 Barg. Corresponding propane temperature is -42 oC, on the assumption that liquid propane still remains in the vessel (conservative ).
Depressurizing through e.g. a METSO valve (http://www.cheresour...chromatography , post 27) results in same final temperature (besides valve gets quite open at end of depressurization).
So my understanding suggests that min operating temperature in flare header should be the normal boiling point of propane. This principle is generally applied here, but other opinions reflecting different practices would be welcomed, as well as comments on this text.

#6 ankur2061

ankur2061

    Gold Member

  • Forum Moderator
  • 2,484 posts

Posted 21 April 2012 - 10:24 AM

kkala,

Please try to understand what controlled depressurization means?

In controlled depressurization the time of depressurization is so adjusted that the flare system (header and downstream) temperature is not allowed to fall below the limit of the minimum design metal temperature. The orifice is sized accordingly.

Ambient conditions do play a role in preventing the metal temperature to fall below the MDMT by exchange of heat from the warmer ambient temperature to the colder pipe temperature during a controlled depressurization. The controlled depressurization ensures that the temperature downstream of the orifice does not fall rapidly enough to fall below the MDMT of the pipe metal.

The combination of the rate of depressurization and the exchange of heat between ambient air and the pipe metal thus ensures that the minimum temperature seen by the pipe does not fall below the MDMT.

Well I hope this explanation should clear all your misunderstandings.

Regards,
Ankur.

#7 kkala

kkala

    Gold Member

  • Banned
  • PipPipPipPipPip
  • 1,939 posts

Posted 21 April 2012 - 02:45 PM

Thanks, though (mis)understanding remains. In the propane example, depressurization should stop at 0.76 barg to avoid temperatures lower than -29 oC (how?). Ball valve is assumed always opened.
Accepting min operating temperature=-42 oC, min design temperature would be -42 oC (-47 oC per Norsok standard P-001, 4.3.2), process-specified. This is considered in post No 3. There may be a chance for temperature downstream orifice never to approach -42 oC, I understand, but the minimum temperature has to be proved (in fact we would conservatively consider -42 oC).
Note: For CET and MDMT see http://en.wikipedia...al_temperature and http://www.eng-tips...cfm?qid=129904 . That is another cup of tea.

Edited by kkala, 21 April 2012 - 03:05 PM.


#8 Butterfly

Butterfly

    Gold Member

  • Members
  • 73 posts

Posted 25 April 2012 - 08:57 AM

Thank you for all your explanations. I will go into a deeper thinking with the propane issue (I am, by the way, dealing with a refrigeration cycle with propane)

#9 kkala

kkala

    Gold Member

  • Banned
  • PipPipPipPipPip
  • 1,939 posts

Posted 27 April 2012 - 04:13 PM

It is noted that minimum design temperature of depressurization systems and piping from deethanizer or C3/C4 splitter to flare header connection (described in http://www.cheresour...-pcv-on-spheres '> http://www.cheresour...-pcv-on-spheres , post No 27) has been process specified as -42 oC.
Also note that web reference in post No 5 is erroneous, correct reference is the one written in this post.
Main flare header has min design temperature probably - 5 oC. But this was designed in 1985, when no LPG depressurization took place and LPG PSVs were discharging to atmosphere.

#10 ankur2061

ankur2061

    Gold Member

  • Forum Moderator
  • 2,484 posts

Posted 29 April 2012 - 11:39 AM

All Design Engineers,

Presently I am studying a report on depressurization of natural gas from well flowlines. The report mentions that the calculated vapor outlet temperature downstream of the RO is -70°C which is below the design temperature of LTCS of -46°C and which would not allow LTCS to be used under such circumstances. The depressurization studies were conducted using HYSYS dynamic depressurization utility.

It is also well know that the HYSYS heat transfer correlations are not accurate enough and using HYSYS in such a case would invariably give lower temperatures forcing the use of alloy steels such as Duplex Stainless Steel for the pipe downstream of the restriction orifice.

OLGA simulator also has capabilities to perform depressurization studies and it goes one step further in accurately defining the boundary layer between the vapor temperature and the inner wall temperature of the pipe. The report mentions that when they used OLGA for the depressurization studies, a minimum vapor temperature of -70°C resulted in an inner wall temperature of -46°C of the pipe downstream of the orifice due to heat transfer effects of the pipe metal with the outside surroundings. The report concludes that on the basis of this study done in OLGA it has been decided to retain LTCS for all the depressurization lines where the vapor temperature has been calculated to be -70°C or above.

This is for the information of all those who are involved regularly in doing depressurization studies using HYSYS Dynamic Depressurization utility. Remember there is a major cost difference between LTCS and DSS. You will be doing the client a great service if you can help him save the cost of switching from LTCS to DSS based on accurately evaluating what the pipe metal can see as the lowest temperature during a depressurization operation.

Regards,
Ankur.

Edited by ankur2061, 29 April 2012 - 11:39 AM.


#11 kkala

kkala

    Gold Member

  • Banned
  • PipPipPipPipPip
  • 1,939 posts

Posted 29 April 2012 - 06:14 PM

1. According to post No 7 (bottom) minimum developed temperature has to be proved. If this can be realized by a reliable calculation, the result has to be accepted. Of course there are more conservative approaches (post No 9), and everybody can judge the specific case and take responsibility.
2. I would be reluctant to accept -46 oC metal temperature versus -70 oC bulk gas temperature (metal temperature would be considered as -70 oC), yet I am not familiar with such conditions and reported software. It may be realistic, but this can be assessed after much investigation and preferably after a HAZOP study. Advice from members having already faced such cases would be appreciated.
3. In previously mentioned example of propane depressurization, I understand that pressure and temperature upstream orifice are in equilibrium (saturation). Heat gain will increase propane discharge, but at end of depressurization mentioned temperature will approach -42 oC as pressure will approach that of flare header (if there is still liquid propane).
4. Post No 10 has introduced the potential possibility of a design metal temperature higher than bulk vapor temperature, of course if there are substantial savings from the material without practical "discount" in safety. Some further data from the natural gas depressurization report could help understanding.
(α) Bulk gas temp = -70 oC, inner wall temp = -46 oC, outer wall temp = ? oC, ambient air temp = ? oC, wind velocity = ? m/s (to see conditions downstream orifice, pipe assumed on air).
(β) Pipe diameter? Temperature e.g. 50 or 100 m downstream from orifice?
(γ) Actual gas temperature is invariably higher than what HYSYS calculated.
5. Minimum design temperature is specified in Norsok standard P-001 (4.3.2). For depressurizing propane minimum design temperature is undestood as -47 oC, but advice is welcomed. "Temperature calculations shall as a minimum include heat transfer between fluid and vessel" is not clear to me.

#12 ankur2061

ankur2061

    Gold Member

  • Forum Moderator
  • 2,484 posts

Posted 29 April 2012 - 11:12 PM

2. I would be reluctant to accept -46 oC metal temperature versus -70 oC bulk gas temperature (metal temperature would be considered as -70 oC), yet I am not familiar with such conditions and reported software. It may be realistic, but this can be assessed after much investigation and preferably after a HAZOP study. Advice from members having already faced such cases would be appreciated.
4. Post No 10 has introduced the potential possibility of a design metal temperature higher than bulk vapor temperature, of course if there are substantial savings from the material without practical "discount" in safety. Some further data from the natural gas depressurization report could help understanding.
(α) Bulk gas temp = -70 oC, inner wall temp = -46 oC, outer wall temp = ? oC, ambient air temp = ? oC, wind velocity = ? m/s (to see conditions downstream orifice, pipe assumed on air).
(β) Pipe diameter? Temperature e.g. 50 or 100 m downstream from orifice?
(γ) Actual gas temperature is invariably higher than what HYSYS calculated.


kkala,

I know for sure that you are not familiar with either HYSYS or OLGA which are industry standards in simulation and modeling in upstream oli & gas. I am not obliged to present the entire report on the forum. Your reluctance to accept -46°C as the inner wall temperature is due to the fact that you have never done such calculations before and would neither understand them and obviously what one does not undertstand, one would be reluctant to accept.

Another point is how does HAZOP help identifying the minimum pipe inner wall temperature. The minimum pipe inner wall temperature is a calculated value. It appears that wherever you don't undertstand the design implication you invariably put in the word HAZOP to confuse the readers.

That heat transfer does take place between the pipe metal and the ambient is something that cannot be denied. Since the whole process is very dynamic, steady state heat transfer calculations cannot be applied and only rigorous simulators such as HYSYS and OLGA can perform such calculations. When comparing the heat transfer correlations of HYSYS with OLGA for dynamic heat transfer the OLGA correlations are found to be more realistic.

As you can see post # 10 was addressed to design engineers and not for forum writers with a fertilizer operations background.

Edited by ankur2061, 29 April 2012 - 11:54 PM.


#13 fallah

fallah

    Gold Member

  • ChE Plus Subscriber
  • 4,659 posts

Posted 30 April 2012 - 01:10 AM

It is also well know that the HYSYS heat transfer correlations are not accurate enough and using HYSYS in such a case would invariably give lower temperatures forcing the use of alloy steels such as Duplex Stainless Steel for the pipe downstream of the restriction orifice.


Dears,

IMO, the main issue of using HYSYS for depressurizing study will result in such conservatism in MDT determination mentioned by Ankur relates to "Isentropic Efficiency". It allows using up to 100% for vessels just containing gas while all know it wouldn't be accepted in real world.

Anyway, an usual procedure in working with HYSYS is: starting with 100% isentropic efficiency which would give lowest temperature and then if there would be no low temperature issue regarding the economic material selection no further work is required, otherwise in order to save the material cost a lower efficiency (for which an agreement with client to be prepared) can be used based on previous experiences and engineering practices.

Fallah

Edited by fallah, 30 April 2012 - 01:11 AM.


#14 kkala

kkala

    Gold Member

  • Banned
  • PipPipPipPipPip
  • 1,939 posts

Posted 30 April 2012 - 05:21 AM

Concerning post No 12 by ankur2061, I would expect specific answers to the simple questions posed instead of generalizations. The limited data asked in para 4 of post No 11 is not the entire report. As said, -42 oC has been locally applied as minimum design temperature for LPG depressurization. One has to be conservative for something unknown and knew. Requested figures are not enough, but could help a bit.
Post 12 is full of hints that are not in favor of whom writing them. I do not think they degrade me. Moreover views in an engineering forum should have been supported through data, not impressions.

#15 ankur2061

ankur2061

    Gold Member

  • Forum Moderator
  • 2,484 posts

Posted 30 April 2012 - 06:20 AM

kkala,

Is it in the same manner that you refused to provide the simple answer to the Mobil standard reference regarding the valve closure time of >5 minutes for pipeline surge analysis?

And you still have not answered the question that how HAZOP helps in determining the minimum metal temperature during a depressurization.

Please provide backup for reference of "locally" applied minimum design temperature of -42°C for LPG depressurization. And even if such a "local" practice is there, has anyone done a dynamic depressurization calculation to determine or verify the pipe inner wall temperature.

As far as data is concerned following is the data:

Min. Ambient temperature = 0°C

Wind Velocity = Max: 37.5 m/s; Avg: 5 m/s

Outer Wall temperature is a calculated variable and changes very rapidly along with the depressurization profile, increasing even more rapidly then the inner wall tempertature. The minimum inner wall temperature is of concern in deciding the MOC of the pipe i.e. LTCS or stainless steel.

Well now that you have been given the data, I do expect that you will provide me the depressurization calculations to prove that -46°C as the inner wall temperture is not correct which as mentioned in your post # 11, you are reluctant to accept.

#16 kkala

kkala

    Gold Member

  • Banned
  • PipPipPipPipPip
  • 1,939 posts

Posted 30 April 2012 - 06:56 AM

Post No 13 by fallah points out a weak point of HYSYS, presenting lower than actual temperatures in depressurization. Following points are noted.
1. The main question is a bit different: to safely accept metal pipe temperature of e.g. -46 oC, versus a real bulk gas temperature of -70 oC ("a minimum vapor temperature of -70°C resulted in an inner wall temperature of -46°C of the pipe downstream of the orifice").
We have specified -42 oC as minimum design pipe temperature for LPG, either upstream or downstream depressurization (post No 9). We have so far generally specified minimum design metal temperature equal to minimum developed process temperature (not higher metal temp, both -42 oC in the example). This may be conservative (post No 7). Specific examples of specified process temperature versus min design metal temperature would be appreciated; to see possible differences (if any). Is it supported by some practice or code?
2. Mentioned principle of min design metal temperature has been applied for at least 20 years, and not only locally. So the issue may be of interest for a lot of colleagues. It is true that there are still old main flare headers that have higher min design metal temperature (post No 9).
Could an inspector be convinced that all safety matters are satisfactory in case of a relevant accident on these old headers? Steels get fragile at low temperatures, but I it is not easy for me to have a risk feeling.
3. As said in post No 11, liquid propane depressurization is expected to have min design metal temperature = - 42 oC upstream depressurization, and probably higher downstream (how much?) according to the methodology taking heat gain into account. Is it so? I understand post No 13 speaks of the flow downstream depressurization (throttling through orifice / valve is not isentropic) and problem appears only when developed process temperatures go below -46 oC, limit of killed carbon steel (post No 10).

#17 kkala

kkala

    Gold Member

  • Banned
  • PipPipPipPipPip
  • 1,939 posts

Posted 30 April 2012 - 07:35 AM

Concerning post No 15 by ankur2060, data of the simulation is understood to be:
Bulk gas temp = -70 oC (*), inner wall temp = -46 oC (*), outer wall temp = variable ?, ambient air temp = 0 oC, wind velocity = 5 m/s (red figures are given now).
-42 oC is normal boiling point for propane, look also at post No 11 para 3.
For -46 oC versus -70 oC, please read previous posts.
Style of post No 15 again not proper, bye-bye.

(*) just downstream orifice

Edited by kkala, 30 April 2012 - 07:43 AM.


#18 ankur2061

ankur2061

    Gold Member

  • Forum Moderator
  • 2,484 posts

Posted 30 April 2012 - 08:01 AM

Concerning post No 15 by ankur2060, data of the simulation is understood to be:
Bulk gas temp = -70 oC (*), inner wall temp = -46 oC (*), outer wall temp = variable ?, ambient air temp = 0 oC, wind velocity = 5 m/s (red figures are given now).
(*) just downstream orifice


kkala,

The data has been provided to help you understand. Does your understanding now conclude that -46°C pipe inner wall temperature is not correct and if yes, then could you please calculate and provide the actual pipe inner wall temperature based on your calculations. Looking forward to your calculations on dynamic depressurization.

#19 kkala

kkala

    Gold Member

  • Banned
  • PipPipPipPipPip
  • 1,939 posts

Posted 30 April 2012 - 09:47 AM

The data has been provided to help you understand. Does your understanding now conclude that -46°C pipe inner wall temperature is not correct and if yes, then could you please calculate and provide the actual pipe inner wall temperature based on your calculations. Looking forward to your calculations on dynamic depressurization.

Good joke! Please read post No 14 carefully, as well as last sentence of post No 17.

#20 ankur2061

ankur2061

    Gold Member

  • Forum Moderator
  • 2,484 posts

Posted 30 April 2012 - 12:09 PM

Good joke! Please read post No 14 carefully, as well as last sentence of post No 17



Yes a good joke for a joker who has no idea of what dynamic depressurization is all about.

#21 kkala

kkala

    Gold Member

  • Banned
  • PipPipPipPipPip
  • 1,939 posts

Posted 01 May 2012 - 04:24 PM

Question of post 18 was "misleading". In the spirit of last para of post 14, readers are kindly requested to try to clarify this important matter (if they know about it). Previous posts contain several not answered questions, more could be asked on the fly.
Note: HAZOP study here (after basic engineering) could solve the issue, since a HAZOP leader would be hired with broad exprerience on the specific subjects (informed about in advance), able to know applicable cases and risks.
This does not concern the topic, but replies to a point of post No 12.
Hopefully forum will clarify the matter, bye-bye from me.

Edited by kkala, 01 May 2012 - 04:43 PM.


#22 ankur2061

ankur2061

    Gold Member

  • Forum Moderator
  • 2,484 posts

Posted 01 May 2012 - 11:26 PM

Note: HAZOP study here (after basic engineering) could solve the issue, since a HAZOP leader would be hired with broad exprerience on the specific subjects (informed about in advance), able to know applicable cases and risks.
This does not concern the topic, but replies to a point of post No 12.


Wow! This is news for me that during HAZOP you try to determine the minimum wall temperature in a depressurization. I always understood that a HAZOP review is done subsequent to a Design Review and a design review is the correct forum to address issues related to design including design temperature. Well that can be understood, since if a person has never attended a design review he would have no idea what I am talking about.

As far as HAZOP is concerned a HAZOP facilitator or leader conducts the HAZOP using the HAZOP methodology adopted. A HAZOP leader should be well versed in the various HAZOP methodologies and should have a broad understanding of the overall process. A HAZOP facilitator should also be involving in node identifciation on the P&IDS under HAZOP review. It is not necessary for the HAZOP leader to be conversant with design issues related to the project. In fact. a preamble for HAZOP is that during HAZOP there will be no design review since it has already been conducted and closed. I had written a blog entry on HAZOP - "Processs Guidelines on Conducting a HAZOP review" where a MS-Word attachment has been provided and which has been liked by many.

http://www.cheresour...a-hazop-review/

I will not be writing anymore on this post since it is like trying to make a small child understand his follies, foolishly hoping that the child will learn. But then I keep forgetting that age has nothing to do with childishness. Or is it????

#23 kkala

kkala

    Gold Member

  • Banned
  • PipPipPipPipPip
  • 1,939 posts

Posted 02 May 2012 - 09:23 AM

Following are notes on post No 22 by ankur2060.
-Topic is to select material in a propane depressurization system.
-Min design metal temperature of -42 oC has been locally applied (upstream and downstream depressurization).
-Another view supports a temperature higer than -42 oC (how much?). A simulation example reported in post No 10 calculates local minimum metal wall temperature = -46 oC, versus local gas bulk temperature = -70 oC, under ambient air of 0 oC and wind of 5 m/s.
-Probably the example would not change material for propane, but it might save money for other cases of lower temperatures, if acceptable. Replying to questions of posts No 11 and No 16 (and others derived by readers) would help for this, expelling some vaguness, though not completely.
-Technical replies, in addition to two red numbers above (wind not clear), are needed. Of course this is not obligatory, but members could help limit incompleteness of this topic.
- A HAZOP leader can be just coodinator, or be hired through stricter criteria. The latter was locally applied in the past, concerning two LPG refinery installations. But the topic is indifferent to HAZOP affair, being a response to post No 10, not a cause for argumentation.

Edited by kkala, 02 May 2012 - 09:39 AM.


#24 Propacket

Propacket

    Gold Member

  • ChE Plus Subscriber
  • 260 posts

Posted 04 May 2012 - 08:11 AM

No technical response to this topic. Just a few notes.

I am a process engineer with >3 years experience in oil and gas sector and too young to teach people who have lived their entire life in their profession. I seek apology from my senior members for my comments i am going to write.

To be frank, professional ethics in this thread have been annihilated. Someone is calling another a joker and a child. I am a great fan of Ankur's posts/blogs/spreadsheets. But its very disappointing to see your mentor calling another person a joker. Yes there may be some misleading, confusing and aggravating comments by kkala. However, it is unfair to call him a child/a joker and making fun of his knowledge about depressurization or anything else. If someone (in your opinion) is reluctant to accept, you should leave him as he is. Remember, being reluctant to accept is part of one's personality and you cant change it. If you cant change something, it is better to leave it not to break it. You people are the persons we learn from and whom we see as mentors. I hope clarification will come on my remarks. Otherwise, I was just frustrated to see those comments and wanted to oust my frustration.

Sorry,

Regards,

Haseeb Ali

Edited by P.Engr, 04 May 2012 - 08:32 AM.


#25 kkala

kkala

    Gold Member

  • Banned
  • PipPipPipPipPip
  • 1,939 posts

Posted 20 July 2012 - 10:37 AM

Having expected responses to above post No 24 for 2.5 months to no avail, I have to congratulate Haseeb Ali (P.Engr) for the courage of his opinion. One should not necessarily be senior or elder to "teach".
Cold blood and independent judgement are advantages in debates / confrontations to approach a right solution. Insulting phrases do not help this purpose. Intolerance would have eliminated them, but in a real world this does not seem easy.
Although undesirable, engineers can face similar complex events in their careers (with real consequences), so some familiarization of such cases may be useful.
Regarding the potentially "misleading, confusing and aggravating comments by kkala", one has to point out the limited available time and that clarity (especially in English) is not easy, concerning the query or the reply. Cases can require a long answer to cover the query aspects, real or assumed, and patient reading to "catch" them. By specific questions / comments you can help improving the answer in technical aspects or verbal clarity (which could improve whole personal style, too).

Edited by kkala, 20 July 2012 - 11:05 AM.





Similar Topics