21 replies to this topic

[Propacket]

Hi Guys,

We have a gas HCDP (HC dew point control unit) consisting of a gas/gas exchanger, a propane chiller and a low temperature separator. Operating pressure of low temperature separator is 600 psig and operating temperature is -35 degF. Minimum ambient temp is 32 deg F. We have performed a detailed blow down study of the system taking into account all the relevant factors (heat input from ambient, metal mass in contact with vapors). Outcome of the study is that minimum fluid temperature in the vessel can drop to -80 deg F. However, minimum metal temp in contact with the fluid is -20 degF due to heat input from hotter ambient. So we concluded that low temp carbon steel (-45 deg MDMT) is adequate materiel of construction for this service. Now i need to know the minimum time before repressurization after blow down of the system has been performed. For example, if the blow down takes place inadvetantly by operator, Can we start the plant as soon as we come to know ?

[flarenuf]

hi

You have to consider two asepcts here on represuring the vessel

1) liquid Ko or hydrate formation on initial fill of the vessel which will be sitting cold, you say the residual liquid will be at -80F and the vessel -20F. Based on the tempertures you are quoting gradually repressuring the vessel should not be a problem. If you have access to Hysys Dynamics i would try a simulation in there to satisfy yourself.

2) yield stress(rupture stress) of the vessel at T and P. you may want to read
http://www.eng-tips.....cfm?qid=116637

attached doc may be of help
flare - e-nuf

Attached Files

•  Scandpower - Guidelines Pressurised Systems exposed to fires 2004.pdf   827.97KB   93 downloads

[fallah]

P-Engr,

What is the MAWP of the separator? If the mentioned MDMT (-45, not specifying the unit but it may be deg of centigrade) is specified at separator's MAWP?

Fallah

[flarenuf]

fallah
Low temperature Carbon Steel is good dowb to -45C

The MAWP of the vessel is its design pressure and not related to teh low temp.
flare-e-nuf

[fallah]

flarenuf,

MDMT relates to MAWP and you can lower the MDMT with lowering MAWP.

Fallah

[Bobby Strain]

P-Engr,
I would expect that the vessel is insulated and the minimum metal temperature is closer to the fluid temperature than you indicate. If I was an owner or designer, I would select materials appropriate for the lowest fluid temperature. If it is an existing system, I would avoid any depressuring without first draining the liquid. Shortcuts like these lead to accidents.

Good luck,
Bobby

[Propacket]

flarenuf:

How can i model repressurization in Hysys?

[Propacket]

Fallah

Actually, we are checking whether low temp carbon steel is adequate if depressuring is performed. What we do know from the vessel data sheet is -45 degC MDMT. Not known at what pressure it is reported. I have already mentioned that outcome of the study is that minimum fluid temperature in the vessel can drop to -80 deg F. However, minimum metal temp in contact with the fluid is -20 degF due to heat input from hotter ambient. So we concluded that low temp carbon steel (-45 deg MDMT) is adequate materiel of construction for this service. Will this information work?

Edited by Art Montemayor, 07 August 2012 - 05:59 PM.

[Propacket]

Bobby Strain:

Selecting materials based on the lowest fluid temperature is highly conservative approach. By the way we have considered insulation in calculating heat input from ambient.

[fallah]

P.Engr,

Use UCS-66 of the ASME BPVC SEC VIII DIV I, to determine the lowest MDMT of the material you are using (LTCS). But considering the conditions of your case, seems there would be no problem in the sevice you described.

Fallah

[Propacket]

Fallah:

Okay. What about my original question "repressurization"?

[Bobby Strain]

P.Engr,
I still believe that you missed some points. You must determine minimum metal temperature at the coldest location. This includes the nozzles located on the bottom of the vessel. I expect that, since the liquid is boiling, the local metal temperature will approach the liquid temperature. The problem with the carbon steel is that it becomes quite brittle at low temperature, and can break like glass. So you need to think through your whole procedure.

Bobby

Edited by Bobby Strain, 07 August 2012 - 05:07 PM.

[Propacket]

Bobby Strain:

Hi
You are right. I was only looking at the vapor temperatures. I overlooked the liquid temperatures. Yes, in contrast to the vapor, metal temperature at the bottom (-64 deg F) is quite close to the liquid temperature (-73 deg F). So you think we should base our material selection on this temperature?

[fallah]

P.Engr,

As a general rule you should consider that: When depressurizing a vessel will result in temperatures lower than the MDMT for full pressure, repressurization shall not be done until the vessel warms up to the minimum design temperature for full pressure.

Fallah

[Propacket]

fallah:

And if depressurizing a vessel results in temperatures higher than the MDMT for full pressure, repressurization can be done without any hesitation. Got it. Thanks a lot fallah.

Edited by Art Montemayor, 14 October 2012 - 10:20 AM.

[Propacket]

Bobby Strain:

Hi
You are right. I was only looking at the vapyor temperatures. I overlooked the liquid temperatures. Yes, in contrast to the vapor, metal temperature at the bottom (-64 deg F) is quite close to the liquid temperature (-73 deg F). So you think we should base our material selection on this temperature?

Fallah,
Can you look at this issue as well. Should we base our material selection on the bottom liquid temperatures?

[kkala]

More than two months have passed without a response on MDMT of the propane gas - liquid separator / vessel. Below is an opinion, comments could help for a clearer picture.
1. Vessel operating pressure is 600 psig and temperature -35 oF (post 1).
2. Vessel will be insulated (post 9).
3. Lowest fluid temperature is not always a conservative approach for material selection (post 9), especially if the fluid is liquid. We have locally applied it even for gas, discussed in http://www.cheresources.com/invision/topic/15093-designing-flare-systems-for-low-temperature ^b. Liquids have higher heat capacity per unit volume (specific heat x density), meaning inertia to temperature change due to heat losses. Also higher partial heat transfer coefficient to metal wall, meaning close temperature between liquid and metal.
4. All steel mass of vessel has apparently one minimum wall metal temperature (MWMT).
5. During depressurization, fluid (gas) temperature goes down to -80 or -73 oC, but calculated heat input from ambient (32 oF) causes a temperature of -20 oF on the contacting vessel wall; liquid temperature goes down to -73 oC and (calculated) temperature on the contacting metal wall (vessel bottom) is - 64 oF (posts 1, 8,13).
Not having experience on this, it seems to me that something has escaped from calculated results. Probably the study considered vessel steel mass to arrive at -20 / -64 oF, but then one uniform steel temperature would have been estimated. Considering above paras, I would assume -80 oF as minimum temperature of vessel steel, after having checked it again as minimum fluid temperature developed during depressurization.
6. Mentioned minimum steel temperature would occur at lower pressure, probably atmospheric, not at vessel design (or even normal operating) pressure. But MDMT seems to depend only on steel wall thickness (given for a specific vessel) and steel plate qualification, irrespectively of actual pressure (ASME VIII div1, UCS-66); see http://www.cheresources.com/invision/topic/16048-difference-between-sa-515-and-sa-516-pressure-vessel-plates/, post 5.
7. So required MWMT has to be -80 oF, according to the above (80 oF to be rechecked).
8. Reported MWMT of the vessel is -45 oC = - 49 oF (post 8)
If temperatures lower than MWMT can occur, the vessel needs to be impact tested.
There are also other possibilities to eliminate development of low temperatures, e.g. inserting gas (N2 or other) at a pressure during emptying the vessel.

Edited by kkala, 14 October 2012 - 09:51 AM.

[Propacket]

Thanks kkala, I will review your comments and follow up.

[Propacket]

Not having experience on this, it seems to me that something has escaped from calculated results. Probably the study considered vessel steel mass to arrive at -20 / -64 oF, but then one uniform steel temperature would have been estimated. Considering above paras, I would assume -80 oF as minimum temperature of vessel steel, after having checked it again as minimum fluid temperature developed during depressurization.

The depressuring was carried out reach atmospheric pressure in blow down time corresponding to the valve size obtained from fire case depressuring. Yes, theoretically, all the steel should be at one temperature but this is not so in Hysys depressuring results. In my opinion, Hysys does not take into account the heat transfer due to conduction within the steel mass from high temp zone to low temp zone (from metal mass contacting gas to metal mass contacting liquid). That's why two different temperatures are reported.

Mentioned minimum steel temperature would occur at lower pressure, probably atmospheric, not at vessel design (or even normal operating) pressure. But MDMT seems to depend only on steel wall thickness (given for a specific vessel) and steel plate qualification, irrespectively of actual pressure (ASME VIII div1, UCS-66);

I disagree with your opinion that MDMT does not depend upon pressure. -49 degF MDMT i mentioned above is at vessel MAWP and it decreases as pressure decreases. So When i calculated MDMT reduction due to decrease in pressure to atmospheric pressure, resulting MDMT was much lower than the original value of -49 degF even lower than the lowest fluid temperature. For that reason, we finally selected a MDMT of -49 degF. The only problem is that vessel should not be repressurized until vessel gets heated to reach -49 degF.

Edited by P.Engr, 16 October 2012 - 09:12 AM.

[Fritzy]

There is a lot of great comment in the posts above. However the question remains unanswered. Upon repressurisation of an isolated system the mass of gas entering the system is performing work on the gas that is already within the system by compressing it. The effect is that upon repressurisation of the system the gas temperature will rise above the inlet gas temperature. Locally at the repressurisation point there will be a temperature drop according to the pressure differential however overall in the isolated section the temperature will rise. I have attached a relevant section from F J Moody's "Unsteady Thermofluid Mechanics" to help understand the situation.

 Repressurization.pdf   59.65KB   72 downloads

Regards,
Fritz Gruber

[Robert Montoya]

Hello, the equipments must be in normal level operation, the filling should done little by little to avoid a sudden expansion occurs. Once the accumulators/vessel are in their normal operating value, start the compressors.
Note that the final pressure in this case should be at least 100 psi; during the filling from the pressure up to 600 psi, the temperature decrease may be important, this procedure should be analogous to that done during the start up of the plant .

[kkala]

The depressuring was carried out reach atmospheric pressure in blow down time corresponding to the valve size obtained from fire case depressuring. Yes, theoretically, all the steel should be at one temperature but this is not so in Hysys depressuring results. In my opinion, Hysys does not take into account the heat transfer due to conduction within the steel mass from high temp zone to low temp zone (from metal mass contacting gas to metal mass contacting liquid). That's why two different temperatures are reported......
I disagree with your opinion that MDMT does not depend upon pressure. -49 degF MDMT i mentioned above is at vessel MAWP and it decreases as pressure decreases. So When i calculated MDMT reduction due to decrease in pressure to atmospheric pressure, resulting MDMT was much lower than the original value of -49 degF even lower than the lowest fluid temperature. For that reason, we finally selected a MDMT of -49 degF. The only problem is that vessel should not be repressurized until vessel gets heated to reach -49 degF.

1. It is safe and logical to assume liquid and vapor in equilibrium and one uniform temperature resulting from depressurization, e.g. -80 oF or whatever calculation will indicate. It is pointed out that there is liquid in the insulated vessel. Any heat input from ambient air has negligible effect.
It is also noted that the case differs to a bare gas pipe exposed to air. Looking into http://www.cheresources.com/invision/topic/3504-isentropic-efficiency-and-its-implications and http://www.cheresources.com/invision/topic/15342-minimum-design-metal-temperature may be worth while.
2. Concerning dependence of MDMT on pressure, please look into http://65.57.255.42/ubbthreads/ubbthreads.php?ubb=showflat&Number=7357, saying that MDMT depends on the material and its nominal thickness. Of course thickness depends on design pressure; but for a given vessel, wall thickness is what is. So MDMT seems to be same for (say) maximum pressure or half of it. This can mean that an impact crack will not save the vessel, if it operates at low pressure. This was the base of saying that MDMT depends on wall thickness, not on pressure. Reasons supporting that MDMT depends on pressure are welcomed.
Above link on MDMT can be found in above post no 17 (by kkala), second link. Question on MDMT versus thickness has been placed to Forum there (no response so far).
Meanwhile mater was further examined. Look at above link on isentropic efficiency, posts No 2-6. In a few cases reduction of pressure seems to lower MDMT. Studying UCS-66 of ASME Sec VIII div 1, as well as ASME fig 323.2.2B, could clarify the matter.

Edited by kkala, 26 November 2012 - 06:17 PM.