Vacuum Protection on Pressure Vessel 3

[Tonylmiller]

I am working on a relief system for a Section VIII vessel rated at 60 psi MAWP. I am familiar with overpressure protection, but I am wondering about vacuum protection on this vessel.

If a scenario exists that could produce a vacuum, do I need to provide a vent for that? The U1 report for the vessel does not specify a vacuum rating.

I have API RP 520 (both parts) and 521, and API 2000 but I do not see this addressed. Do I need another document?

Thanks in advance,

Tony

Tony Miller
Cornerstone Engineering Group, p.c.

http://www.cegpc.com

 

[TD2K]

If a credible scenario is going to result in a vacuum created that is greater than the vessel's ability to withstand you either need to engineer out it out, handle it via procedures (practical only in specific cases) or provide vacuum relief.

If the U-1 form doesn't specify a vacuum rating you can always calculate it. Vacuum capacity for a vessel can be provided by adding external stiffening rings, we did this recently for a 10' ID tower. For some scenarios, such as forming a vacuum during steam-out, some companies rely on procedures to ensure adequate venting is provided.

Vacuum protection can be provided by introducing a non-condensible gas into the vessel (nitrogen, fuel gas, etc). Air can be used if your process will tolerate air AND if leakage of your process fluid through the air inbreathing device during operation doesn't create other hazards. Unfortunately, because of the uncertainity that could result in a vacuum, there's no standard vacuum breaking requirements similar to those listed in API 2000 (which I'll note may still may not cover you for all possible storage tank inbreathing cases)

 

[jte]

Well put, TD2K.

Tony-

In determining the vacuum rating for the vessel, you may also be able to take advantage of tray support rings if you're dealing with a trayed column with single pass trays. There are restrictions, though, so read the code (VIII-1 UG-29(c))yourself on this.

What kind of material (CS?) diameter, thickness, distance between stiff sections (heads) are we talking about?

jt

 

[reena1957]

Hi,
Many of the recent equipment data sheets I came across, specify Full vacuum along with the positive design pressure and do not rely on any procedure or Vacuum relief valve. I feel such engineered solutions are more reliable than depending on procedures or devices.

 

[TrevorP]

I concur with reena1957. If it's a new vessel, specify full vac. Then you're covered, you don't have to worry about air ingress, or specifying the right size vacuum relief valve etc. If it's an old vessel not rated for full vac., then you need to consider what vacuum rate is required and size accordingly.

Vacuum is often possible due to thermal inbreathing, collapsing of vapours, such as when steam cleaning of vessels, and vessel content reduction, for example when pumping out. It is particularly important to consider start-up and shutdown conditions. An internet search will provide dozens of examples of tanks being sucked in due to inadequate vacuum protection. Sadly, it's not an uncommon problem.

 

[sshep]

Dear All,

With respect to full vacuum rating, a previous thread I started (thread124-105493) asked the question of how much extra a vacuum rating cost. Per Icarus using a base pressure of 400kPa (60psig), it was my finding that the cost adder was less than 20% for both CS and SS tower shells. This is 20% of the shell cost, not total cost. If the shell costs $100K (which can still get you a nice size atmospheric tower shell) I am certain that our sites in the USA could not purchase and install a high volume N2 blanket system for the cost of that extra 20%.

My question was prompted by work at an overseas site which did not specify full vacuum, but instead used large volume N2 vacuum breaking valves and piping, redundent pressure instrumentation, and administrative locking procedures to protect from collapsing towers. Incidently just last week that site was investigating an event where they lost N2 supply and pulled vacuum on one of there non-vacuum rated towers- fortunately they think no damage. However, in one of Roy Sanders books he has a picture of a distillation tower collapsed by vacuum and laying on the ground, this is much less frequent of occurance than a tank collapse, but can obviously happen.

As far as I am concerned, full vacuum is the only way to go. Anything else is risking problems for marginal savings at best.

best wishes, sshep

 

[abeltio]

most oil and chemical companies have standards for the specification of pressure vessels.
such specs usually have a guidance for the determination of the MAWP and if vacuum is possible, calculate for: full vacuum, period.

HTH

saludos.
a.

 

[Homayun]

If it's total condensing overhead system of a tower, then just specify Full Vacuum, end of story. But if it's for steam out, there are different approaches. Some companies, like where I work for, don't design for FV in case of steam out. We just tackle it in procedures. But i also have seen other companies where they do desigh for FV.

 

[MortenA]

Tony

As Homayun notised it may depend on company standards. Shell e.g. wants full vacuum in refinery service where steam out is used.

Best regards

Morten

 

[Tonylmiller]

Thanks for the responses. It is an old vessel and unforunately there is no vacuum rating.

I am at home now, and do not have the vessel's details. . .

I guess I'll recommend some kind of device that will prevent a vacuum.

Tony Miller
Cornerstone Engineering Group, p.c.

http://www.cegpc.com

 

[mrtangent]

Tony,

All though the vessel may not be stamped vacuum rated. I've seen vessels re-rated or rather had the stress calculations peformed for vacuum T/P conditions. It would be worth considering that cost. At our site we have an internal ingrity group which would faciliate this.

We had vessels re-rated to higher pressure as it turned out they could with stand it.

T

 

[BenThayer]

if the vessel is designed for 60 psig, it may be capable of a pretty decent vacuum rating. get that sorted out and then see if there are any true scenarios that can reach it (don't forget about start-up, shutdown and cleaning scenarios thought!).

for typical reactors on site (1000 gal. to 2000 gal.), we see that we can frequently get 50-70 psig if a vessel was originally designed to full vacuum and 30 psig. you're basically trying to go the other way with a 60 psig vessel that could be rated for significant vacuum with an inspection and calc review. even if it is not good for FV, the pumps and drain-out scenarios may not be capable of pulling it to, say, 2 psia and you might decide no credible scenario exists.

the mechanical integrity group and engineers should be your best friends for this type of work!

 

[TD2K]

I would strongly suggest you first consider checking the vessel to see how much vacuum it can withstand.

It avoids the whole problem of vacuum protction, avoids the issues/questions with the reliability of a vacuum breaker system, avoids the need to calculate 'how much gas do I need to supply' which can be a non-trival calculation as well as the cost of the vacuum breaker system.

To me, the upfront engineering cost to evaluate what is needed to provide inherent vacuum protection through design is minor compared to the potential savings.

 

[Tonylmiller]

Thanks for all of the comments. I do have software to calculate the rating, but I discovered that there are no credible scenarios for creating a vacuum. It is an ammonia plant, so all vessels are normally under pressure (not vacuum) but the existing PSV also had vacuum protection.

I guess the next step would have been to calculate the vacuum rating. It might have been o.k., since it is rated for 60 psi positive pressure.

Whew!



Tony Miller
Cornerstone Engineering Group, p.c.

http://www.cegpc.com

 

[BenThayer]

just make sure that there are no scenarios that can cause vacuum during start-up and shutdown. the worst incident i have ever seen was a mixup during a start-up at an ammonia facility.

the start-up sequence required the vessel to be charged with liquid prior to the introduction of a vapor. the operator started adding vapor and realized his mistake and added the liquid while the vapor was still in the vessel.

the vapor was ammonia gas and the liquid was water. the result was a massive cave-in of the vessel.

HAZOP's should define the credible scenarios and we engineer solutions to prevent them or mitigate them.

ammonia plants with potential for ammonia vapors are not immune to vacuum scenarios.

 

[Montemayor]

Excellent point and well described by Ben!

To my knowledge of 45 yrs in the field, I know of no 100% sure-fire exception to the viable vacuum scenario. While most vacuum scenarios may be discarded as not probable, there is always a possibility - albeit remote and not likely. However, this just proves the point that the best effort an engineer can do is to take a serious look at a detailed Hazop(s), involving all the possible disciplines.

There is nothing quite as tragic as witnessing a young, bright engineer seeing his future career go down into Limbo as one of his vessels "sucks-in". This stops being a joke or funny the moment it affects us. Vacuum is a powerful force that has impacted on many promising engineering careers in the past. Don't let it happen to you. And I'm not even focusing on the severe possible safety hazards involved.

 

[Tonylmiller]

Yes, that is advice well given and well taken. This particular vessel has a vent line that is always open to atmosphere.

Tony Miller
Cornerstone Engineering Group, p.c.

http://www.cegpc.com

 

[Christine74]

Quick rule of thumb:

If a carbon steel vessel's thickness is greater than or equal to its diameter divided by 100, it will calculate for full vacuum per ASME without needing stiffening rings, regardless of its length.

-Christine

 

[MJCronin]

Christine...

Although you are probably accurate with your rule of thumb for cabon steel vessels..... Can you give us a source ???

An engineering book or text carries much more weight when explaining or justifying a design to others...


-MJC