Pressure Vessel Design Critique

Sure, it'll be fine.

Youmreally haven't supplied enough information.

You mention it's a reactor

We do not know what process to make any kind of opinion.

Is this a ASME certified vessel?

Thanks Lyn, firing it up now!

Thanks for the response, it is not certified, and it is a prototype.

Let us know how it works for you.

...if the <...we...> survive, of course...

You do know the Mk. 1 prototype never, ever works, right? Hey, it is a prototype.

Are you making a pressure-cooker bomb? If so, we cannot help you, but your help is already on the way (NSA, ATF, FBI, etc.). If not an intended unit of destruction either personally or otherwise, then maybe we can help you directly.

If you do not have a remote location to undertake these reactions, do not do this. What you are describing has the potential to release suddenly a lot of energy upon your person. Any device designed to withstand such pressure and temperature and containing water falls under the pervue of the State Boiler Inspector where you live, so you need to review the Boiler Code for your State.

You must install a safety relief valve, because even though you state a quoted temperature and pressure, you cannot guarantee that you will not drive the heat above that point, so no, your vessel is not safe for the intended purpose.

Some of your specifications appear to be in order, but I question the use of Viton O-ring on that. Not sure the O-ring will not become semi-plastic and extrude out of its groove. That is even if the metal holds up OK to direct weed burner exposure.

Apparently, there is something wrong with your specification, as I see 250 mm diameter tube at one point then a 300 mm O-ring later on. Fix that. Also, you seem to have a good material for O-ring as long as the pressure-temperature line is not exceeded. The question is one of local overheating of the metal with lack of heat transfer might well cause an O-ring failure.

a flange flairs out?

As funny as your comments have been, can you please provide some insight into the flaws of this design so that they can be rectified?

The only flaw seems to be the blind flange thickness as a PN16 flange has been specified.

I take it that direct to me. Wether it is or isn't, isn't the point.

Flaws? I can't.

Who actually did the design?

Again there is not enough information, on the full welds, what size and procedure.

Were the welders certified that did the welds? Type of filler?, weld prep?

When you hydrotested, what temperature? There still needs info on the orings, being viton is normally a good start..... how deep is the groove for encapsulated?....

As far as the blind flange, there should be enough info for that already, there are rated flanges, 150#, 300#, etc,.... With the size being 16mm? , it may be a problem.

Sorry.

I am not qualified to evaluate, or approve, this design.

Without knowing more about the gland design and O-ring properties noone can say for sure.

The flange is probably OK.

phoenix911, the comment was directed to me.

That's a relief

Thank you anyway for your thoughts.

The blind flange has a groove that has been machined into it, providing a 0.25mm gap between the o-ring and the internal wall of the groove and a 2.4mm gap between the o-ring and the external wall of the groove.

The welders are certified and made the welds much thicker than they normally would. 316 wire was used and phosphoric acid pickling.

Hydrotest hasn't commenced yet as it was discovered that the flange is only rated to ~225psi (PN16). However does this rating only apply if using a full face insertion gasket as opposed to something more effective such as an o-ring? Thanks.

It's been a while since I was doing ASME, welds being thicker? You mean the size of the bead reinforcement? That's not necessary a good thing, one time we did that and our AI (Authorized Inpector) would not approve it. Had to do with our WPS.

with a certified welder? What certifications?, this may not be an issue.... It really too difficult to give any recommendations here.

fyi,

Here is a link about excessive weld reinforcement on a bead that I mentioned earlier, read the paragraph header "Convex Beads"

GA

Were the vessel to be containing more than 215bar-litres of energy when pressurised, then it constitutes a piece of equipment that would fall under the UK's various pressure vessel regulations. Such a piece of equipment requires burst/collapse indemnity insurance assessment, irrespective of its prototype nature. In turn, this would mean submission of the design to the prospective Insurer for assessment prior to construction, and inspection during construction with a hydraulic test by that company's Engineer/Surveyor, who will issue a written report on the vessel should it pass its various tests as a record of the assessment. At that time, the user will subscribe insurance premiums prior to the vessel being used.

• This sort of post is typical from people who are good at welding and have never designed a process vessel before. A bit of pipe. Some flanges. Bounce it off some "mates". That'll be OK. Yeah, right...

No mention has been made of the nature of the process reaction. Exothermic? Endothermic? The wetting agent and the wetted materials? Any prospect of runaways? What about the maximum and minimum temperatures and pressures to be expected? Pressure relief and safe venting arrangements? [Rhetorical questions - NNTR] These are fundamental to the design of any process vessel; the type of bolts used is a rather tertiary consideration!

A Chemical Engineer would look at the process and look at ways of converting it to a continuous process, rather than batch, look at ways of reducing the inventory of hazardous materials within the plant while maintaining throughput, and look at ways of carrying out the process at reduced temperatures and pressures. Only then when the process parameters have been correctly set would the reaction vessel be specified, and only after then would the Mechanical Engineer become involved in the design of the vessel. So in this particular post, the tail is wagging the dog.

One can only hope that the <...we...> are either adequately insured against blunders, or the process vessel is installed far enough away from work colleagues or centres of population to minimise the risk when it explodes; let no CR4 reader suffer as a result...

Neither. It is simply a criterion.

• A car tyre, for example, is too small, and need not be considered.

How are you going to regulate your burner so the stainless steel temperature does not go over the 220°C for the viton o ring. Which most sources list max temperature at 204°C. I would use a stainless steel O ring or some other O ring compound with higher temperature.

As far as the O ring holding 300 psi. Most are rated for 5000 psi in hydraulic applications.

IIRC, Class 300 8" blind flanges are 1-5/8" (1.62") thick, so on a quick view yours don't seem thick enough.

Thanks for the link Phoenix, I didn't realise that thicker welds could be a bad thing but that makes sense now. The flanges are most likely cast and then machined to size, and the welds in this unit were standard fillet welds.

My real question was, will the flange be suitable for an operating pressure of 265psi even though it is only rated to ~225psi as I was told that flange ratings are provided under the assumption of full faced insertion gaskets being used as opposed to more efficient sealing systems like O-rings. I then thought that I may as well discuss the entire design in case there are any clear oversights which it looks like there are. I am an engineer but this is not my field, and the other engineers in this company are cowboys so I am glad that I have made this thread and I will now ensure that the vessel is inspected and approved by a certified engineer.

What other information will I need to provide for a mechanical engineer to analyse the vessel?

Viton is compatible with the fluid, however a small scale test will be conducted after hydrostatic testing to confirm this.

The main query I have with the O-ring design is that the groove dimensions from the manufacturer are specified such that the O-ring sits hard against the ID of the groove. I thought that conventionally, the O-ring should sit hard up against the OD of the groove however I have followed up with this multiple times with the manufacturer and they state that it is fine.

The length of the vessel is 500mm and heated by a burner underneath which covers ~350mm and is centred. The vessel slowly rotates to allow mixing and even heating of the steel. A hood has been fabricated that leaves the flanges exposed to the air but allows the unit to heat up quicker. Internal temp needs to reach 160 degrees Celsius. A temperature gauge within a thermopocket monitors this.

A stainless o-ring may be the way to go, although there is a 60 degree difference in temperature there, surely the internal temperature wouldn't vary that much?

Exactly right, ANSI class 300 flanges are min. 47.7mm thick.. triple the thickness! Hence why I am worried that the current flange won't be able to handle the pressure.

However, a 50mm thick flange would be too hard to for workers to repetitively manually handle when assembling and disassembling.

Perhaps a DN150mm vessel should be used with a PN21 flange (rated to 305psi and 24mm thick)?

Thanks to all so far for the valuable information that you have shared.

I read through your description twice and found several disturbing things. You never mention anything about a pressure relief valve. You are making an enormous assumption then that your heated bomb cannot produce more than 265 psi. A pressure that is 18% over the flange pressure rating you cite. You also do not indicate at all what chemical(?) reaction is desired inside this vessel. The external thermal energy applied is the only energy source you control and that control is only by implication.

If you find a certified engineer willing to review this design, bring them your original vessel fabrication and assembly dimensional documentation. Don't forget to include your system control designs. Don't forget to tell them your operation plans. My point here is if this is a one time only use with remote operation and monitoring possible then you maybe worrying more than it is worth. Try it as is. If something breaks with nothing valuable around, who cares. If this is the crux of a frequently repeated, expensive operation then many other concerns will need to be addressed.

Most importantly, listen and respond to the questions asked by the external engineering firm. This should be an interactive process, not a one time data dump.

End of the 5th paragraph, in the OP. says "relief valve".

I was so focused on the reply. Yes, the OP does mention a relief valve. Whew.

Still, how often this vessel will be used and safety mitigation might make this thread "much to do about nothing".

...or "Bhopal the Second". Who knows [rhetorical question - NNTR]?

One wonders where the vented materials might get sent. Perhaps SolarEagle has an amusing illustration - somewhere between a botty-burp and Bhopal? <wink>

Paragraph 2 is not far off describing a formal Hazard and Operability Study (usually abbreviated to HazOp). Do Mechanical Engineers do those?

PWSlack is essentially on the money here. I think the best course of action will be to go with a smaller diameter bore (6") in order to reduce the pressure load on the flange and ensure easier manual handling for workers assembling it whilst specifying much more suitable flanges (PN35 - 507psi). I will get started on drawing up plans for this and get in contact with a local engineering company qualified and experienced in this field to approve the design, inspect and test the unit. We were hoping to achieve ~10 batches with this unit, however we may come back and get this unit retrofitted with acceptable flanges and appraised if the DN150 unit is successful. Thank you all very much for your time.

Hold the horses! Get a Chemical Engineer to look at the process design. First!

Will do, PWSlack - thanks for the constructive criticisms.

Is the overall design that I have specified, i.e. a piece of pipe welded between two flanges like a cotton reel not the standard way that a process vessel would be fabricated from?

Are there any special welding methods or types used on process vessels such as this, so that I have some idea before drawing up anything further? Cheers

I like to chime in here.

If it was a pipe and being used as such there may not be a problem, but its be unsed as a reactor, with modifications such as o-rings, flanges.... and the rating of the flanges seems to be under sized.

As far as welding techniques, full penetration welds, (if the internal welds can be visually inspected that's great, if not, you would need a backer plate on them to insure full penetration.

And if you can see the internal welds for visual inspection, but you can only weld on the outside due to space restrictions, purge the tank when welding.

I do hope you're joking, Mildred. I really do hope you're joking.

Do you intend to get this "thing" assessed for insurance or are you going to sit on it while your reaction is taking place, Mildred?

Flat materials do not resist differential pressure very well, and hence must be quite thick. Rounded shapes are much better, and can be lighter. If you choose cleverly, your materials can probably be 3 mm thick rather than 9 mm.

If your reactor can be less than 6" I.D., it becomes a piping component rather than a pressure vessel, which may eliminate some code issues. Longer and skinnier is your friend.

I don't know if 3A sanitary pipe/tube components go up to 6" (150 mm) in size, but that type of construction could be the most economical choice.

I am looking into domed flanges, however I don't think they would suit due to the different instrumentation required to be welded onto it.

If flat flanges are the only option available, what are the most effective sealing flanges that can be provided from the manufacturer? I.e. raised face, machined groove?

So lengthening the vessel doesn't increase the force that will be applied on the flange as it doesn't increase the surface area for the pressure to be applied to? Does this mean that there is no difference between a 500mm long vessel and a 700mm long vessel at all in regard to the forces acting on the flange?

Welds can be conducted internally and externally and visually inspected afterwards. Pulsed MIG welding was used to weld the flanges together - is this the best method of welding for this application or is something like TIG more appropriate? Thanks,

If forces were the only critical thing to the safe design of a reactor vessel then there would be little need for engineering codes on reactor vessels.

I hope that this is a scholastic homework assignment. If not you will probably hurt somebody someday.

Redfred,

Can you please detail the other main considerations when designing a reactor vessel, as I thought I had covered everything: temperature, pressure, material compatibility with feed material.

With a 316 stainless 150mm reactor, utilising schedule 40 pipe and ANSI class 300 flanges, 3000lb fittings, Viton o-ring with groove machined, the components are all rated above 500psi at 200 degrees Celsius, when the operating temp/pressure is 200 degrees @ 265 psi for 4 hours.

The only concern I now have is the welding types/process.

Operating temp ~170 degrees Celsius.**** not 200.

Can I detail the other main considerations when designing a reactor vessel?

You do realize that I do not get paid for my comments on CR4 and that I consider my personal expertise is in electrical circuit theory...

Is the reaction in the reaction chamber endothermic or exothermic?

If neither then why do you call it a reaction vessel?

If you are not certain that any or all plausible reactants might interact with the vessel and all seals, then why should I?

If you hold your own engineering analysis of what might be critical in such a low regard that you have to ask an anonymous public forum for support, why should I (my anonymous self) proffer you any unpaid advice as more relevant than any other schmuck that wandered upon CR4.

Please, PLEASE, PLEASE!! do not risk anyone's safety on my or anyone's comment here on CR4. Be careful!