Maximum Allow Operating Pressure and Maximum Allowable Working Pressure are terms used when setting PSVs and choosing materials etc. Are these terms identical?
The maximum allowable working pressure is set by the process engineer. This will be totally dependent on the process. The Maximum Allowable Working Pressure is a value calculated by the vessel manufacturer determined after the vessel is physically designed. It takes into account the design pressure and design temperature, both of which are again set by the process engineer.
The MAWP is always referenced to the top of the vessel in its normal operating orientation. By that I mean if it is a vertical, cylincrical vessel, the MAWP refers to that pressure at the top when the vessel has been set in place on its pad.
One point to note is that unless specifically asked for, many, if not most vessel manufacturers will stamp your vessel with your specified design pressure and temperature, not the true MAWP. They will only calculate the MAWP and stamp same if asked to do so and this may, or may not incurr an additional cost.
I think Phil's first sentence has a "typo". I believe his intent was to have it say that the MAOP (Maximum Allowable Operating Pressure) is set by the Process Engineer of record - not the MAWP.
Phil is well above most of us all when it comes to the subject of safety relief - so I know it has to be a typo. I want to make sure no one misinterprets what he meant.
I will demonstrate the difference with an example:
A plant has a process operating at 30 psig, with a normal operating range from 20 to 50 psig - but pressure spikes 20% above normal operating pressures are to be expected; the minimum pressure is 0 psig. 50 psig is the MAOP. They want to expand and require a new vessel - so they go to their engineering group and ask that they spec a vessel for them. The designer of the vessel will prepare a spec for the fabricator, he must select a MAWP for the fabricator to design and construct the vessel to code. This MAWP needs to be a comfortable margin ABOVE normal operation AND the relief device for it needs to be set no higher than the MAWP AND this relief device set pressure needs to be a comfortable margin ABOVE the MAOP plus normal pressure spikes.
The margin of comfort is a variable; certainly the relief device set point chould be at least 10% above normal peak pressure spikes; and the MAWP should be at least 10% above normal peak spikes as well - others will argue 10% is not enough.
TO complete our example using 10% as a minimum margin:
MAOP = 50 psig
MAOP with 20% spike = 1.2 x 50 = 60 psig
Assuming we do want to set the relief valve = MAWP (same as Design Pressure):
60psig/0.9 = 66.7 (say 70 psig).
So the minimum is 70 psig - what is the maximum? The pipe system is rated for 150 psig maximum - so that s the maximum. The designer and the plant consult nad decide that 70 psig is enough margin and they could take a vessel designed for 150 psig but they don't want to spend the extra money.
So they specify a relief device set for 70 psig and spec a vessel for 70 psig at whatever temperature was required.
Last step is the fabricator for the vessel, he requires 0.17" steel plate for an MAWP of 70 psig; but he has 0.25 inch plate he wants to use to clear out his shop. His calcs show 0.25 inch plate is good for 135 psig; but after fabrication he only STAMPS and codes the vessel for 70 psig because the designer did not ask for max rating of the vessel.
So when the plant gets its new vessel they have to accept 70 psig as the MAWP instead of 135 psig which they could have gotten for the same price.
Further, if they had gotten the vessel stamped at 135 psig, and if the rest of the process piping in their plant was good for 150 psig; then they could have set the relief valve at 135 psig and max normal operation would have been that much farther away from the setpoint of the relief valve decreasing further the possibility of a release ever occuring.
Sorry it took so long, but this can get pretty interesting; and heart breaking if you screw it up way back in front-end loading your project.
The margin by which your MAWP must exceed your maximum allowable operating pressure is also in part determined by the properties and type of relief device used. Spec'ing this difference too closely can result in the need to select a very expensive relief device, or in frequent releases from the relief device.
Very true. If I use a rupture disc in front of the relief valve, for example, then the 10% margin jumps to 30% minimum since a tension disc is fatigued above 70% of its rated burst pressure.
I like the example given by CHD01...good work! My only comment is that I reserve the term "MAWP" for the vessel fabricator only. In process design, we only give the fabricator our design conditions. He comes up with the true MAWP after building the beast, as you pointed out.
I would be careful of setting a relief valve at 90 percent of vessel operating pressure. Many types of reliefs (typically those with hard seats) will leak (simmer) at 90 percent of set pressure. A valve with leakage at 90 percent will leave a repair or test shop with a clean bill of health in many applications.
In a case like that you almost always need to install a close tolerance rupture disc under the valve.
I agree, 90% of set was defined as the limit of normal operation of max normal operating pressure plus pressure spikes; at that point and above, leakage can be expected to occur with some installations.
The process designer had better be involved in selecting both the MAOP and the minimum required MAWP. You're correct that the actual MAWP will depend on what the fabricator uses to construct the vessel- but as I mentioned, the selection of MAWP is dependent on MAOP requirements AND relief device design/selection. You design too tight a spread between these two parameters at your peril- your design might require an extremely expensive or unreliable (or both) relief device.
I intentionally did not sugest a recommended ratio between operation and design conditions because there are too many variables that enter in to this selection, instead I simply stated what I consider the absolute minimum. Using the 30% margin I discussed in my previous post is a much better ratio since it allows the use of tension type rupture discs which the 10% margin would not; and the greater this margin is obviously the more flexable the process is for expansion - but the cost increases as well normally. I think the example I gave is clear enough to be of help to the new design engineer. Nothing is absolute.
I'd like to clarify something about the MAWP of an ASME vessel. The MAWP of a vesesl is what the vessel Manufacturer puts on the ASME Data Report.
ASME VIII states that the MAWP is based upon the as-designed features, such as shell thickness, repad size, weld sizes, etc. In this case a set of ASME calcluations is required that eventuaslly identifies the weak link of the entire vessel. In the absence of such efforts, ASME permits the Manufacturer to use the vessel design pressure specified by the Purchaser.
Today there are many canned, simple-minded ASME calc programs that will grind out the calcs and find the weak link for you. Not so, twenty years ago. As a result, many exixting vessels have an MAWP that is merely a reflection of the design spec, not the harder-to-figure potentially higher MAWP.
If I could make a suggestion to the spec writers herein. If you want the highest MAWP possible on the vessel, you have to specify in the contract that you require the Manufacturer to determine the MAWP by calculation of all of the the essential parts of the vessel. Otherwise, you get the easy answer that MAWP = Design Pressure.
I want to personally thank you for this succinct and authoritative information. I, and probably most of the above respondents, have knowledge of this procedure (or lack of it) but it is very valuable information for those that haven't accumulated the scar tissue and hard lessons that come with the trade.
What you have stated is precisely what I have personally experienced in the field. Project engineers, in my opinion, should have a secondary residence address at their favorite fabrication shop - if they take their job seriously.
I have also found this dilemma/problem with respect to API storage tanks - especially, as you stated, when their age reflects the calculation methods of old. This comes up frequently nowadays when young engineers are assigned to upgrade or change service on storage tanks and they are faced with a mandatory Nitrogen blanket that requires a dead pressure span of 10 to 20" of water column and the tank is stamped as "atmospheric". I believe the API stamp, in most cases, reflects analagous results to what you've described.
I, as most of my colleagues, am not looking for an official and proven MAWP just because I like the academic calculations and pure theory. We are expected to design to the MAWP benchmark and it behooves us, as engineers, to know not only that it exists - but to be able to quantify it. Agreeably, in some cases, there are no economics to justify it being identified if the alternative design figure exists and it suffices. However, as in the case of API tanks, the quantification of the MAWP (and the identification of its underpinnings - or weakest points) is of large importance to both process and project engineers with interest in both operability and safety.
Thanks for bringing up the fact that the MAWP is based on the weakest link for all the vessel components versus the design pressure the fabricator is given, I should have included that in the example to be complete.
