Vapor recovery unit sizing 1

[TD2K]

I'm working on a job that involves retrofitting NG gas blankets and vapor recovery units (VRUs) onto some existing tanks.

I have the flow rates and the fluid properties of the materials that I can run down to the tanks.

As backup for the VRUs, the tanks will be equipped with vacuum/breather valves sized per API 2000.

The question I have is wrt the VRUs. The tanks vary in size from 25,000 to 150,000 bbls. Maximum flow rate to each tank is about the same and the estimation of the displaced vapor flow rate to the VRU is straight forward. However, the thermal outbreathing requirements are substantial and are significant greater (up to 7 times greater for the largest tank) than the gas displaced by liquid movement into the tank. I’ve talked to the VRU vendor we are likely going to go with as well as the tank breather valve technical group (this person also has a good knowledge of VRUs).

The VRU vendor has quite a few units in operation and in some cases, the customer has sized the units and in other cases they have sized the units. They size the units on the basis of the gas displaced by the incoming liquid and do not include the thermal outbreathing requirement, their customers have used the same basis. None of their units have had a problem with meeting capacity and not venting excess gas to the atmosphere through the vent valves. They recognize the discrepancy but believe the fact their units do not vent suggests the API 2000 thermal outbreathing flow rates are very conservative. The person I talked to at the tank breather valve company said that in his experience, some companies do include the thermal outbreathing requirements in the VRU capacity basis but this is a company practice as opposed to a process or fluid property that might result in higher capacity requirements.

For example, a 150,000 bbl tank in high flash point service (greater than 100F) requires a thermal outbreathing capacity of 47,500 scfh. This is equivalent to raising the internal temperature of the gas, assuming the tank was intially empty, by about 30F which seems high to me to actually achieve. The tanks will normally contain product which will be above ambient temperature and I was wondering if this might moderate the internal pad gas temperature such that the API 2000 thermal outbreathing requirements are effectively ‘smoothed’ out.

I’d appreciate any comments/experience others have. Sorry for the long post.

 

[EGT01]

Let's see who can create the longest post.

My experience is limited so I'll be interested to see what others say too but I have dealt with a hydrocarbon vapor recovery system some years back under the direction of a more senior engineer and just recently had to design a 36wt% muriatic acid storage vent system.

What I learned from the senior engineer was to account for the outbreathing and which seems reasonable. For a given tank, as the sun rises and starts heating the tank, the vapor in the tank will try to expand causing tank pressure to increase. If you are pumping in at the same time, the vapor space is compressed further increasing the pressure in the tank. It would seem if you account for only the vapor flow due to one effect, you won't be able to control the tank pressure.

Ok, but how are you accounting for the vapor flow due to each effect? I'm not quite sure I followed your comment "They recognize the discrepancy but believe the fact their units do not vent suggests the API 2000 thermal outbreathing flow rates are very conservative."

First, I don't think it is mentioned in API 2000 but if you go to their web site for API 2000 technical interpretations,

http://www.api.org/techinq/

you will see that API does not recommend using their thermal venting rates for sizing vapor recovery systems since it may result in excessive capacity that will rarely, if ever, be used. From API 2000, thermal venting rates are based on a 100F/hr change in temperature.

Now if you are using the API 2000 suggested values for venting capacity due to liquid movement into a tank, there may be some conservatism there. The high flash stock number (about 8.57 scfh/gpm) is only slightly greater than the actual liquid displacement (about 8.02 cfh/gpm). The low flash stock vent rate is double (about 17.14 scfh/gpm). However, look at the Appendix in API 2000 which explains the idea about allowing for more than just liquid displacement such as accounting for some evaporation but how much evaporates may depend on whether the liquid is introduced through the vapor space of the tank and the liquid properties.

As for ways to account for the "normal" thermal outbreathing, my mentor had suggested 5F/hr. I think you could relate that to a typical ambient temperature change of say 70F to 95F over a 5 hour period (morning to noon).

Also, consider what the EPA method for emission estimates (AP-42) uses. Seems I remember that they use a 20F average diurnal temperature change for accounting for thermal breathing losses. However, I don't think they count thermal breathing losses for storage tanks that are insulated by assuming the vapor space is unaffected by ambient conditions. If you're unfamiliar with AP-42 go here

http://www.epa.gov/ttn/chief/

and follow the link to Emission Factors.

As for the HCl tank vent system we recently installed, its a venturi/packed column scrubber that I was liberal with trying to account for liquid movement into the tank by using more than the actual liquid displacement to account for the fuming of the 36wt% acid and added thermal outbreathing due to 5F/hr temp change. Considering the position of the storage tank pressure control, the vent system has excess capacity but Operations likes this better than the last system which was marginal to undersized. Of course, there were other problems with the old scrubber system that the replacement addressed so it's hard to say that any one thing resulted in the improved operation.

 

[TD2K]

I should have specified in my original post that the thermal outbreathing requirements I mentioned (4th paragraph) are the values per API 2000 rather than say some process based calculated thermal outbreathing requirement.

 

[EGT01]

Wanted to try to make a quick comment about product temp greater than ambient and effects on outbreathing. From the ideal gas law
V2 = V1 * T2/T1
you can estimate change in volume as
V2 - V1 = (V1*T2/T1) - (V1*T1/T1) = V1*(T2-T1)/T1
the venting rate is (V2-V1)/time where the rate of temperature change is (T2-T1)/time
more rigorously dV = V1*(dT/T)
for further reference on the more rigorous approach see "Esitmate Tank Breathing Loss", Hydrocarbon Processing, Jan 1977.

The units of temperature are absolute so that tends to diminish the effect of product temp above ambient but obviously as T1 increases, V2 - V1 will decrease.

My original post could be misunderstood in regards to the temperature values used. A 20F to 25F change may be a typical ambient temp change between day/night but the tank vapor may not actually experience that and could be at a temp other than ambient. For a more rigorous evaluation of T2-T1, you may want to use something other than a typical ambient rate change. Again, you may want to look at the EPA AP-42 to see how they address this.

In your case, with product temp > ambient, you may actually have a rate of vapor temp change less than the ambient rate of change so your "normal" outbreathing for recovery system sizing may be smoothed out. None of this has any impact on the API 2000 outbreathing requirements that would be used for overpressure protection design. As further stated in the API 2000 Technical Inquiries, "the vent size recommended in API 2000 should not be reduced for tanks equipped with vapor recovery systems."