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

Logic Behind Interlock / Trip


This topic has been archived.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

#1 ankur2061

ankur2061

    Gold Member

  • Forum Moderator
  • 2,484 posts

Posted 07 May 2014 - 01:25 AM

Dear All,

 

I am seeing a interlock / trip scheme which I am not convinced about. Maybe it is because I have not been exposed to such a specific scenario. The background is as follows:

 

A CDU connects to an air-cooled condenser. The air-cooled condenser outlet connects to a reflux drum. The reflux drum is used as buffer for reflux pumped to the crude column. The reflux drum vapor line is hooked to the flare system with a split pressure control loop using Nitrogen. The ultimate protection is with a relief valve sized for fire case. The reflux drum level control is from product rundown. An interlock / trip is provided for the reflux drum level high-high whereby the upstream air-cooled condenser is required to trip the condenser fans.

 

Does this logic make sense? I have not encountered any such scheme and would like to know from other members if they have encountered such a scheme.

 

My understanding is that the reflux drum should have a relief valve which needs to be checked for both fire case and liquid over-fill case to determine the governing scenario and size the relief valve accordingly. Alternatively, a separate relief valve other than the fire case vapor relief valve needs to be provided for liquid over-fill in the drum which will be connected to a separate liquid blow down system.

 

The above is based on the consideration that the reflux drum liquid over-fill is a credible scenario.

 

Let me hear some opinion on this from the learned members of the forum.

 

Regards,

Ankur

 

 

 

 



#2 fallah

fallah

    Gold Member

  • ChE Plus Subscriber
  • 4,930 posts

Posted 07 May 2014 - 01:55 AM

Ankur,

 

In general you are right; but normally we can take credit for operator response following an HHL alarm provided that there would be adequate vapor space inside the drum between HHLL and liquid filled situation.  Then in most similar cases by providing adequate vapor space, just fire case would be considered as only credible scenario for drum protection against overpressure.



#3 prengr

prengr

    Junior Member

  • Members
  • 29 posts

Posted 07 May 2014 - 02:59 AM

Dear Ankur,

 

For your reference, API 521(2008) is descriving for Over-fill case as below.

 

5.23 Overfilling process or surge vessel
5.23.1 General
Many process or surge vessels, including columns and towers, have a liquid level present during normal, start-up,
or shutdown conditions. Experience has shown that this equipment can be overfilled under certain conditions. If
the source pressure of a liquid feed or supply line can exceed the relief device set pressure and/or the design
pressure of the equipment, then overfilling shall be included in the system design.

 

What i understand is that we don't need to consider PSV against Over-fill case of reflux drum. And, if we should consider over-fill case for reflux drum, we can take credit as Mr. Fallah mentioned.

 

In addition, as you explained, if you install a sperated PSV for over-fill case and the outlet line of that PSV is connected to a separate liquid blow down system, it will make another problem because we cannot select the appropriate PSV during over pressure situation. For example, if over pressure is occured due to non condensible gas, PSVs will be opened and then non condensable gas will go to flare header and liquid blow down system. In this case, we can protect distillation equipment against over pressure, but, liquid blow down system might be over pressurised......

 

Anyway, in my point of view, a separate PSV against overfill case of reflux drum is not required.

 

Thanks,

Hugh.



#4 ankur2061

ankur2061

    Gold Member

  • Forum Moderator
  • 2,484 posts

Posted 07 May 2014 - 03:04 AM

Fallah,

 

Thanks for your reposne with regards to the relief valve but my main question is related to the tripping of the air-coooled condenser fans on reflux drum level high-high.

 

Does this trip / interlock make sense? My understanding is that the fan trip of the condenser would cause vapor flooding of the entire system and increased column top pressure. Probabaly I am not seeing it in the right perspective of a CDU operation and safeguard due to my limited experience on such systems.

 

I would like to have some more understanding before I can conclude.

 

Some further information:

 

The CDU column in this particualr case has a PSV directly mounted on it. Also the overhead line from the column to the condenser and from the condenser to the reflux drum has absolutely no isolation valve. The column top sectiom, overhead line, air-cooled condenser and the reflux drum have the same design pressure which is logical. The separation distance from the column to to the reflux drum is quite small (read same fire zone)

 

Now the question is why would I require a  PSV on the refllux drum in the above mentioned scenario. Now with the amendments in ASME Section VIII pressure vessel code, you may choose not to provide a relief valve on an equipment designed as per ASME Section VIII if you can provide a proper justification for excluding such a relief valve.

 

Regards,

Ankur.


Edited by ankur2061, 07 May 2014 - 03:16 AM.


#5 fallah

fallah

    Gold Member

  • ChE Plus Subscriber
  • 4,930 posts

Posted 07 May 2014 - 04:31 AM

Ankur,

 

Thanks for additional info.

I think considering no isolation valve along the column outlet to reflux drum, to stop further build up of the level in the drum it could be reasonable to trip the air cooled condenser on HHLL situation. If the air cooler fans wouldn't be tripped on HHLL, after a while the condenser would be flooded with liquid leading to loss of condenser duty while its relevant fans are still active. Then no matter the condenser is flooded with vapor (fans off) or liquid (fans on), the important thing is loss of condenser duty.

Obviously when the level builds up inside the drum, with or without air cooler fans trip, the column would be pressurized that would certainly have its own sefeguards against overpressure. Of course, to do the exact analysis the whole system should be investigated integrally and with these limited info it might one cannot conclude about the matter anymore.

 

Two equipment with no isolation valve inbetween can be protected with one PSV (or PSV set) provided that the PSV would cover all relief loads of credible overpressure scenarios. The disadvantage of such facility might be mandating of considering the pressure drop effect of moving fluid from equipment without PSV toward one with PSV in relieving situation, on PSV pressure setting. It may lead to decreasing the PSV set pressure less than equipment design pressure that limits the operating range of the equipment. You can check and analyse your case from such standpoint.

 

Sorry for lengthy post...


Edited by fallah, 07 May 2014 - 04:35 AM.


#6 Zauberberg

Zauberberg

    Gold Member

  • ChE Plus Subscriber
  • 2,727 posts

Posted 07 May 2014 - 05:28 AM

Tripping overhead fin-fans will cause a massive pressure relief from such a large system (CDU Tower), because the tower remains heat-underbalanced after the fans are switched off. This by itself shows there was no clear logic on how this interlock could have been considered.

 

On the other issue (LAHH in the reflux drum), one needs to see what are the possible causes of this scenario. These could be:

- Tripping of the product pump

- Discharge LCV/FCV valve failed closed

- Discharge line blocked elsewhere

 

Somebody should look at all these causes, see how credible they are, and try to mitigate overfilling of the drum by implementing other measures, if possible.

 

If LAHH in the reflux drum is a serious concern, then I would trip the feed to the Unit (or tower), rather than switching off the fans.



#7 ankur2061

ankur2061

    Gold Member

  • Forum Moderator
  • 2,484 posts

Posted 07 May 2014 - 11:16 AM

Zauberberg,

 

I fully agree with your observation. Trip of the condenser fan on reflux drum level high-high defies any proper logic. I agree with the logic that the Feed to the column should be cut-off on reflux drum level high-high.

 

I would also propose that the reflux drum pumps should be lined up for auto-start and a logic could be provided for the standby pump to startup on reflux drum level high-high with both pumps running for a short duration to bring down the level in the reflux drum.

 

Hopefully I will have some more comments from senior members of the community regarding their understanding on this subject.

 

Regards,

Ankur.



#8 fallah

fallah

    Gold Member

  • ChE Plus Subscriber
  • 4,930 posts

Posted 08 May 2014 - 02:20 AM

Ankur,

 

Yes, trip of the condenser fans on reflux drum HHLL not to be a good and logical engineering practice but anyway if to be done (the case of yours should be analysed carefully to find the reason as i mentioned) would prevent further level build up and considering following typical safeguards against level build up as a consequence of, let say, inadvertent closing of the suction valve in one of the reflux pumps would create no major consequence:

 

- High pressure alarm(s) by the pressure transmitters on the column or overhead line,

 

- An interlock provided to start the stand by pump on LL flow at the reflux pumps discharge line,

 

- Pressure control provided by a PV on the column or overhead line would releave the excess pressure toward the flare,

 

- Relief valve provided on the column to releave the overpressure due to such consequence,

 

- HL and HHL alarms by the level transmitters on the reflux drum.

 

Indeed, considering such typical safeguards cut-off the feed to the column on HHLL in reflux drum couldn't be a reasonable and normal practice for that situation.



#9 Padmakar Katre

Padmakar Katre

    Gold Member

  • ChE Plus Subscriber
  • 992 posts

Posted 09 May 2014 - 01:34 AM

Hello,

- Typically reflux drum Hi-Liquid level will need intervention of a field operator based on panel personnel i.e. to start standby pump if the level still starts rising. At Hi-Hi liquid level, auto start of standby pump makes some sense.

- What can lead to high liquid level in reflux drum is what Zauberberg as already pointed out. A). Tripping of running pump: start up of stand by pump is action to be followed. B.) LCV / FCV fail: bypass valve opening and reduction in unit throughput is must here if bypass system sizing basis for a lower flow. C) Blockage in D/S system: These product pumps feed the Gas Con section surge drum so unclear about this reason.

-Tripping of fans will lead lower condenser duty (natural draft credit + trim coolers if any), this may lead column over pressurization and liquid carryover to compressor suction drum (very high vapor load), if the carryover is excessive than surge vessel hi-hi liquid level compressor would trip and further pressurization of entire loop.

 

Looking at comments by Zauberberg and Fallah and Ankur's own assessment, this control system should not be opted.

I have worked on crude unit designs/revamps in past but never come across such interlock/trip but only an alarm.

 

P.S.: Ankur, is it an existing facility or a new design. If the system is existing interested to know the operational experience due to such event if any.


Edited by Padmakar, 09 May 2014 - 01:35 AM.


#10 ankur2061

ankur2061

    Gold Member

  • Forum Moderator
  • 2,484 posts

Posted 09 May 2014 - 02:57 AM

Padmakar,

 

Fortunately this is a new design and yet to be implemented so there is an option to get rid of this trip interlock which somebody thoghtlessly decided to incorporate in the design.

 

I do know of certain other units (not a CDU) where the reflux pumps from a distillation system have been hooked up to emergency or uninterrupted power to protect the system from overpressure that can occur on loss of reflux..

 

Your mentioning of operator intervention seems to be quite logical and probably will require to be incorporated in the plant operating manual.

 

I was hoping to have some insight from Art Montemayor.

 

Regards,

Ankur..


Edited by ankur2061, 09 May 2014 - 02:59 AM.


#11 chemsac2

chemsac2

    Gold Member

  • ChE Plus Subscriber
  • 119 posts

Posted 09 May 2014 - 11:32 AM

Ankur,

 

In one of recent projects, I have seen a control scheme that opens pressure control valve to flare in event of high-high liquid level in reflux drum. Reason behind this scheme was to avoid shutdown of the column when reflux drum does not need extra liquid.

 

If column overhead vapour is allowed to condense, in short period reflux drum would be liquid full leading to column overpressure anyway on condenser flooding.

 

Considering air coolers are placed on pipe racks, liquid static head on reflux drums can easily be in excess of 10-20 m. With column PSVs relieving, pressure in reflux drum would be higher by this static head. 

 

Since in your post, you state design pressure of column and reflux drum is same, that trip logic is designers way to deal with this extra static head.

 

Even if reflux drum is designed for this static head, few problematic issues need to be sorted out as below:

  • PSV set pressure for reflux drum i.e. shall it account for static head of liquid for overfill case or keep it lower for fire case relieving vapour
  • Liquid relief and subsequent problems in flare system i.e. flare supports not designed for liquid, flare KOD capacity and flare KOD pump size issues etc

 

In one fractionator in hydrocarcker/FCC, these were the issues we faced.

 

Just my 2 cents.

 

Regards,

 

Sachin 



#12 Art Montemayor

Art Montemayor

    Gold Member

  • Admin
  • 5,780 posts

Posted 12 May 2014 - 09:01 AM

I apologize for being late in responding to this very interesting and important topic.  I believe we have already had some of the top brains within our Forums already contributing to this thread and my contribution is not very different from what has already been discussed.

 

I would not relieve excess liquid level in the reflux drum with a PSV.  I would, instead, relieve the excess liquid level in the reflux drum with a level control & dump valve to a safe, closed drain if there is sufficient pressure or head to drive it.

 

If the PSV on the crude column top head is sufficient to protect both the column and the reflux drum, it should suffice in the event of an over-pressure.  I assume -  as fallah - that there is no way to block the overhead line + the air-cooled condenser + the liquid line to the reflux drum and that the capacity of the PSV is not reduced for the worse case scenario.

 

I would ensure that the source (cause) of the excess liquid reflux in the reflux drum be mitigated first.  This could be one - or a combination - of the following:

  • excess reboil heat;
  • reflux pump failure - electrical or mechanical;
  • reflux control valve(s) fail in closed position;
  • reflux block valves erroneously closed.

I would never shut down or cut off my protection against an over-pressure and subsequent flare release: the overhead condenser.  This is not a cause of the excess liquid in the reflux drum.   I would do the following - or a combination of the following:

  1. Stop all reboil heat ( or reduce it considerably) with an automatic block on the source or an automatic reduction of heat - depending on the ability to react with operator remedial action;
  2. Automatically have a standby reflux pump startup upon the High level in the reflux drum (I identify the high level as the point of remedial action and the High-High level as the point of relief action) - as Ankur mentions;
  3. Automatically activate the LC dump valve upon reaching the High-High level in the reflux drum;
  4. The automatic dump can be done back to the crude column sump if the reflux drum is located with sufficient head or height; if not, the dump should go to a closed drain for recovery;
  5. The crude feed to the column should be activated to the blocked position once the High-High level is reached in the reflux drum and reboil heat is blocked or reduced.

There should be no vapor release to the flare or activation of the PSVs in order to bring the column down to a safe, operable condition where the reflux accumulation problem can be resolved and the column started up again.

 

Additionally, the size (liquid inventory) of the reflux drum should be carefully strapped and identified vis-à-vis the crude column's sump capacity.  In other words, if the column is being fed on flow control, any excess liquid in the reflux drum represents a level drop in the column sump - and an effect on the reboiling ability.  This should be studied and identified as to the effect on the scenario.  The time required for operator reaction to the scenario should be identified with regards to the capacity of the reflux drum to store and dump its inventory.  The capacities and normal residence times in the column sump and reflux drum should be calculated and identified.  (In fact, I would add this data to the P&ID for the operator's attention.)  Certainly, the mechanical design and size of the reflux drum should incorporate some residence time to allow for operator reaction to a reflux failure.



#13 Neelakantan

Neelakantan

    Gold Member

  • Members
  • 124 posts

Posted 13 May 2014 - 09:45 AM

Thanks ART,

 

especially pointing out that the column liquid "load" matrix is essential feature to watch; 

(aside: is there a flow indicator in the feed line?)

 

a very good thread and neat posts by all people; if I can trouble you guys, may be a sketch from Ankur indicating the proposed design and from ART showing a secondary dumpline with dump valve to the closed drain  would have neatly completed the thread, though practicing engineers can visualise them.

 

regards

neelakantan



#14 paulhorth

paulhorth

    Gold Member

  • ChE Plus Subscriber
  • 396 posts

Posted 13 May 2014 - 12:53 PM

I missed the start of this discussion, and I don't really have much to add to the very useful points made, but I would like to make a few comments.

 

The case of reflux drum high high level seems to me to be an upset condition that can happen for any condensing column.  For the various columns in my recent experience (mainly NGL fractionation) I'm pretty sure that this upset would cause a unit shutdown, that is, feed flow and reboiler heat would be shut off. I have never heard of tripping the condenser fans - as this would rapidly lead to high pressure (as others have pointed out) which in turn would cause a unit shutdown. If the cause of high level is reflux pump failure then that also would lead to high pressure.

I agree therefore that this action is not logical.

 

I recognise that shutting down a crude column is a bigger deal than a depropaniser, but I think it would be unavoidable (assuming that operator intervention has failed, as we should).

If Ankur can retrieve the HAZOP report or HAZOP action sheets for this column, I suspect he might find the origin of this strange trip in a poor HAZOP response.

 

On the question of overpressure of the reflux drum by flooding and liquid static head, I would argue that if there is a fire case PSV on the drum set at the same pressure as the main PSV on the column, then, in the event of overfilling leading to overpressure in the column, yes, this PSV would open and relieve liquid to the flare - but the quantity of liquid would be small, just the contents of the condenser outlet line.  Once that liquid had gone, the PSV would relieve vapour together with the PSV on the column itself. I think that's acceptable.

 

A good HAZOP review would bring out all these points, but they are often too rushed.

 

Paul



#15 engg

engg

    Gold Member

  • Members
  • 57 posts

Posted 25 June 2014 - 05:06 AM

how about starting a full reflux  to cool column and reduce feed at same time?






Similar Topics