Tank Vapor Vent Flow Rate Calculations

[OilBoiler]

Hey everyone,

I'm working on designing a vapor recovery system for three API 650 resid tanks that are nitrogen-blanketed. There are two existing tanks that have a system already and I'm checking my design basis against the existing one (they are the same exact tanks) and the flowrates differ greatly, that's why I wanted to check with all of you if I'm doing it right. All the tanks should have the same regime of filling and emptying. These tanks I'm looking at already have vacuum vents that address any inbreathing/outbreathing requirements, but I'm currently calculating the maximum vapor rate I will be having whenever the tanks are filling. I'm assuming that the gases are pretty much 100% nitrogen (is that a correct assumption, since they are nitrogen-blanketed?)This is how I'm calculating it:

1. With the tank levels available through the DCS, I'm calculating the volume rate of the gas displaced whenever the tanks are filled from one specific level to another using the ideal gas law. For instance, the level goes from 4 ft to 30 ft in 3 days, I get a certain volumetric rate of gas displaced in ACFH.
2. From the actual volumetric flowrate, I use the ideal gas law and get lbmoles/hr.
3. Then I convert it to SCFH.

In order to establish my design basis, I look at the trends and by going through the same calculation process, I decide what are the minimum, normal, and maximum vapor flows.

Is this how we calculate tank vapor flowrates based on liquid movement? Because this is resid, with a very high molecular weight and very low vapor pressure, the design for the other two existing tanks assumed that there was no thermal venting. I appreciate very much your feedback. Thank you!

 

[Montemayor]

Oilboiler:

1. Your written description is confusing, so bear with my assumptions. You say you have 3 API 650 tanks filling with Residual Oil; “There are two existing tanks that have a system already and I'm checking my design basis against the existing one” – I guess you’re trying to say: two tanks are in service now and I’m introducing a similar one to work in unison with them. Is that right?
2. What do you mean when you say “All the tanks should have the same regime of filling and emptying”? Do you mean to say that all 3 tanks are filled and emptied in the same manner – from the same fill line and emptied into the same discharge line? Are they pumped out by the same pump(s)?
3. If your Residual Oil is very viscous and has negligible vapor pressure, I believe you are reasonable in assuming that the vapor space in the tanks is essentially 99% pure Nitrogen – that is, IF you don’t have any infiltration of atmospheric air through your vacuum relief valve. Since you haven’t divulged the settings of your conservation vent(s) for over-pressure and vacuum, I can’t be sure if you wouldn’t have any infiltration of air (which would introduce Oxygen).
4. The tank fill scenario should suffice, together with out-breathing allowances, for the normal rate of vapor recovery. However, there are other scenarios to consider: You might have a failed-open scenario on your Nitrogen fill valve and you must allow for this either in recovery or in the design of your conservation vent.
5. Your method of identifying the vapor displacement rate is not what I would accept as accurate or practical. What is done is that the maximum flow capacity of all pumps connected to the tank are summed up in the event all pumps are used to fill simultaneously. This maximum liquid fill capacity is the design displacement rate used to size your recovery compressor or Nitrogen recovery system. If this is a credible event, it should be taken into consideration in designing the recovery system. This is the normal way to do it, together with a capacity control system on your recovery compressor (if you are using one). You haven’t stated how you propose to recover the N2, so I can’t add anything further on this.
6. The total maximum liquid fill rate volume + the maximum out-breathing rate is the total maximum design rate you should base yourself on. Anything less is handled through capacity control. The liquid rate is the actual volume displaced at the tank’s pressure and temperature. Use Uconeer software (available for Free) and calculate the equivalent N2 flowrate.

I have an Excel workbook that I prepared for this type of tank operations, coupled with pressure and relief scenarios. I have offered this in past threads but am not able to furnish my email address any longer. Sorry.

I hope these comments help you out in designing a safe installation.

 

[roker]

hello,

the way you describe your calculations is OK, what about pumping from the tanks Nitrogen is entering the tanks this will complete the sizing of the breather valve.

regards

 

[JStephen]

For thermal inbreathing, you would still have some effect from sudden cooling of the vapor in the top of the tank, even if nothing condensed as a result of the cooling.

 

[OilBoiler]

Art,

Sorry for the confusing description. We are not retrofitting into the existing vapor recovery system for the other two tanks (We have five tanks in total). The existing system for those two tanks will remain independent and I'm designing a system from scratch for the other three tanks.

Thanks again.

 

[Montemayor]

OilBoiler:

Thanks for the feedback. Now the picture is clearer.

I will emphasize once again, for your sake, that the best thing (& usually the controlling factor) is to take the credible scenario of the N2 feed valve failure into consideration. I hope you have already considered this possibility and have ensured that your tank relief system is capable of responding to this credible incident.

You are, I presume, doing at least one or two Hazops on this process change - even though you already have a similar N2 recovery system working on the existing 2 tanks.

 

[OilBoiler]

Thanks for your feedback!

The N2 regulator failure has already been considered for sizing the pressure vents that already exist at the tanks.

I need to check my calculation for design flowrate, because I have second thoughts: The maximum flow I can have either in or out of the tanks is 910 gpm. I converted this to ft3/min and I get 122 ft3/min. I used this as my actual volume rate at which the gas will be displaced. Using this as my ACFM, with the ideal gas law I found the molar flow rate at the tank pressure and temperature. The tank pressure is a few inches of water, so I guess we convert this to absolute pressure to be used in the ideal gas law? According to this I have:

lbmoles/min = ((0.04+14.7 psi)*122 ft3/min)/((10.73 psi ft3/lbmole R)*(385 F+460 R)) = 0.2 lbmole/min

I convert it to standard volume = 0.2 lbmole/min*379 ft3/lbmole = 75 SCFM (for each tank)

Is this correct? The design for the other two tanks that already have the system has a design flowrate of 250 SCFM for each tank and they used the same fill rate I'm using, that's why either my calculations are wrong or they oversized the current system. Our current system doesn't work really well, since it's a water scrubber of vapors and after the vapors are scrubbed, we release them to the atmosphere, however, we still have odor problems. I don't know if the fact that the system may be oversized can be a problem or again, I'm miscalculating the basis.

Again, thank you for all your help!!

 

[Montemayor]

Oilboiler:

You are correct. There are 122 Actual ft3/min exiting the tank (I assume there is only one tank submitted to the 910 gpm inflow). I also notice that the tank(s) is at 385 oF and the outflow must take this into consideration as you have done. You have correctly converted this molar flowrate to “standard” volumetric flowrate. You should have to recover 75 Scfm (@ 60 oF & 14.696 psia).

Normally the system’s ability to handle the vapor recovered should be over-sized and the difference is controlled, as I previously stated, by capacity control – which can be on-off type or valve/pocket unloaders. Whether you do this is largely determined by the economics of the project and the safety criteria that is applied. If your N2 is expensive, you may not opt to vent out any excess venting through your conservation vent. You obviously don’t have to consider thermal outbreathing since you are operating with very hot residual oil at 385 oF – from what I can deduce. One effect, if your tanks are not insulated, that can upset things a little is the fact that the vapor space will cool down during a rain shower. When this occurs, the tank’s vapor space pressure will decrease due to the N2 blanket cooling down. As this is occurring, the N2 feed valve is opening since it is trying to maintain a preset pressure on the tank’s vapor space. That means that this mechanism feeds an excess of N2 moles into the vapor space. Should the residual oil remain hot or not has nothing to do with the fact that now you have initially more moles of N2 in the vapor space than what you calculated at 385 oF. That means that on the next subsequent recovery step you will be pushing out more Scfm of Nitrogen. This may be a reason for increasing the capacity of your recovery system.

Hope these thoughts are of some help.

 

[PAN]

OilBoiler,
Do you use API 2000 to estimate nitrogen consumption for blanketing system? It seems that tank volume is a key factor beside filling and emptying flow rate.

 

[OilBoiler]

PAN,

All five tanks: the two that already have the vapor recovery system and the other three I'm designing a new system for, have nitrogen blanketing already. I have the design basis they used for calculating the nitrogen needed for blanketing and it refers back to Table 2 in API 2000. All of these five tanks have 200,000 bbls of capacity each. I think Table 2 only goes up to 180,000 bbls. The appendix states that tanks greather than the ones available in the table should be considered separately. For the sake of checking the basis I extrapolated the data and I still don't get as high of nitrogen needed as in the design basis.

The nitrogen blanketing system was designed a long time ago (80s) when these tanks were installed so I don't know how they came up with the basis and how they justified as high of flow. For the two tanks, they have nitrogen flowmeters and it indicates an average of 4500 SCFD with very low variation, while the pressure vents and vacuum breakers were designed for 33,000 SCFD. It seems according to the nitrogen flowmeter that the N2 flowrate is independent of fill rate/emptying rate (tank levels), which I didn't expect. Does that make sense?