7 replies to this topic

[DB Shah]

We are in process of installing storage tanks (fixed roof). During detail calculation of N2 PCV capacity sizing (API 2000) the combined requirement of N2 for all the tanks works out to be @ 2500 NM3/hr. The N2 will be tapped from our existing N2 utility header operating at 2 Barg.

I envisage following scenario-
During sudden drop of ambient temperature N2 requirement will be high. At this point all the new tanks as well as existing storage tanks will also tend to draw N2 from the header. This may reduce the N2 pressure in the header drastically and create unsafe condition for all the tanks. (As breather valve will also be installed practically tank pressure will be maintained by sucking air from atmosphere).

Solution thought of -
We also have high pressure N2 in our complex (@ 70 Barg). I intend to put a buffer N2 vessel for the new tanks. This tank will store N2 at 7 BarA, which will be tapped from high pressure N2 header via PCV. I do not intend to store N2 above 8 BarA as the existing utility N2 system has design press 8 BarA.

Problem- Determination of vessel capacity
New tanks are of 3000 / 3000 / 4000 M3 capacity. I have considered 50% average volume in all the tanks this leaves with vapour space of 50% = 0.5*(3000+3000+4000) = 5000 M3.
Prevailing ambient temp = 40° (313°K)
Bulk Temperature reduction due to rain 5°C
Final temperature 35°C (308°)
N2 required = volume of the vapour space reduction =5000 * (1 – 308/313) = 80 M3 at atm press
N2 storage pressure = 7 BarA
N2 storage volume = 1/7 * 80 = 11.4 ~ 12 M3 buffer tank.

Kindly comment if this calculation/assumption is correct OR am I wrong somewhere. (Ideally these tanks should have been floating roofs, but now at advanced stage of project it is not possible)

[proinwv]

First I would not know if the assumptions that you have made are realistic.

Then I suggest that you cannot utilize all of the stored N2 in your buffer. You are assuming that you can by using the 1/7 ratio. You will need to look at the PRV you select to deliver the gas and see its capacity at various inlet conditions and then decide how far down that you can drop the tank pressure and still deliver gas at the necessary flow rate. This will determine the lowest tank pressure that you can tolerate and then you can go from there to establish a ratio to find the necessary volume to store.

Be aware that a PRV's performance is different at different inlet pressures so your volumetric delivery as well as delivery pressure will vary as the storage pressure decays.

[Art Montemayor]

DB Shah:

I am having difficulties in understanding and following what you are describing. You fail to tell us specifically the QUANTITY of storage tanks, the capacity of each, and how they are configured with respect to operation and piping. It is very difficult for me to follow your description without a reasonable flow diagram or, better yet, a P&ID of the proposed installation.

For example, you state “New tanks are of 3000 / 3000 / 4000 M³ capacity”, but you fail to be specific and state that there are only three (3) tanks in this scope and that their total individual liquid capacities are 3,000 m³ for two and 4,000 m³ for the third one. Is that what you mean to write?

Additionally, just how are the tank configured? Do all three tanks store the same liquid product? Or are they handling different products? Are they connected by their vents to each other? Are they being filled at the same flow rates or at different and varying flow rates? Are they “true” storage tanks? For example, do you fill them to capacity and they await a planned or anticipated time when they are un loaded to shipping vessels? Or are these tanks serving as feed tanks to another process and thereby subject to being depleted on a steady state basis? All these various conditions and scopes of work lead to the best manner to design a blanketing system for each or for all three of the tanks.

These are very large storage tanks (792,516 gals & 1,056,690 gals) for being fixed roof type. Therefore, this is a serious and formidable problem that you are facing and you must resolve it with the safety and integrity of the tanks as well as the product they contain. Here, I am assuming something that you fail to state: the product is degradable or potentially harmed by exposure to atmospheric air and requires that any breaking of a vacuum by introducing atmospheric air be done as means of last resort. Am I correct?

I believe you are focusing on an erroneous scenario when you state: “During sudden drop of ambient temperature N₂ requirement will be high. At this point all the new tanks as well as existing storage tanks will also tend to draw N2 from the header”. I do not believe that this will be the case. What normally happens when an unexpected cold front moves in quickly over a storage tank area is that you have a reasonable time to react to the anticipated inert gas demand. A storage tank – depending on size and amount of liquid content – will usually lag a couple of hours in reaching such low temperatures that the blanket pressure will start to drop excessively fast. You have several factors working in your favor in this type of scenario:

• the liquid contents are a favored heat sink instead of the blanket gas;
• the blanket gas is cooled by its contact with the cold walls (assuming no insulation) and this takes time because this heat transfer is very weak due to the natural convection currents and the inherent inefficient blanket gas film coefficient at the walls.

You state you also have high pressure N₂ available @ 70 Barg. But you fail to state at WHAT RATE of feed. You should always state ALL of the basic data, especially flow rates.

You plan to put a buffer N₂ vessel for the new tanks but you fail to state how it will function and how it will be connected. What, specifically, do you mean by “buffer”? Is it just an intermediate blanket gas feed tank for the 3 storage tanks? Again, a simple PFD or a P&ID would be a necessity in understanding what you intend to do.

You state you consider having “50% average volume in all the tanks”. What is the basis for this “worst case” scenario for all 3 tanks at the same time? Is this really a possibility? Not knowing the scope of work for the project puts us here in a quandary as to why you propose such dire circumstances. The basis for your design should rest on the actual and factual method of operations that you are designing for.

I recommend that you furnish us with ALL of your basic data and scope of work for the design – as well as reasonable PFD and/or P&ID sketches in order to fully understand and help you on this topic.

Await your reply.

[DB Shah]

Dear Art/Paul,

Sorry if I could not describe the problem in a clear way.
I am attaching an excel file which I think am able to give all the details.

Tank capacities
2 x 4200 M3 Ethanol tanks and one 3400 M3 tanks storing Ethyl acetate.

Each tank have individual vent and no interconnection.

Ethyl Acetate tank is product tank which will recieve ethyl acetate from day tank of plant(same complex) and despatch ethyl acetate in tankers.

Ethanol has seasonal availability, we intend to store ethanol in two tanks during certain period of the year and use it year round. Ethanol is a raw material for ethyl acetate.


I tried to caluclate the tank bulk temperature drop at 5°C LMTD and even with high Ud of 50 kcal/hr-m2-C (considering natural convection)the drop of temperature in the tank is 0.02°C. I have also included this calculation in the file. So what you are suggesting is correct that the scenario seems to be erraneous.

I intend to use existing low pressure N2 available at 2 Barg for normal operating case. During high demand if N2 pressure drops to 1.5 Barg N2 from buffer will be available to the new storage tanks. Up stream pressure of N2 PCV will be specified as 1.5 Barg to the PCV vendors.
In normal period I will make up HP N2 in buffer tank.

Buffer tank will be just an intermediate blanket gas feed tank for the 3 storage tanks.

(You state you consider having “50% average volume in all the tanks)
I have no basis for 50% liquid level. This was just for a starting calculation.

My root problem is restriction of instant draw of such a high N2 flow from existing header. I can spare high pressure N2 during normal operation (year round) and store it in a buffer vessel which will be used in the above scenario.

I think I could have framed the question in a simpler way - API requriement is @ 2500 NM3/hr of N2 flow rate. For how much period of time this flow rate is required? If N2 is required for 5 mins, quantity of N2 required = 2500*5/60 NM3. I think this would have been a simpler way of addressing the issue.


Thanks & Regards
Divyang

Attached Files

•  N2 Buffer.xlsx   25.77KB   169 downloads

[Art Montemayor]

Divyang:

Thank you for submitting your detailed Workbook.

You communicate well and I now fully understand what you are proposing and why. Your workbook has some errors, I believe, and I am working on adding my comments and calculations to it. I need some time, however, so I will submit it for your review after this weekend.

I consider this project as an important challenge for you and I want to make sure I can contribute only information of worth to you based on my past experience. The fluids involved are, in my opinion, worthy of close and carefull scrutiny and safety. I have what I consider to be important recommendations and I will share these with you in the workbook.

Please pardon my delay.

[Art Montemayor]

Divyang:

Thank you for being patient in waiting for my response. I am attaching my copy of your workbook with the additions I have made. I didn’t edit or modify your sketch. However, I do have some comments to add to it.

Basically I checked out your gas calculations because I thought they were resulting in too high an answer. My results are detailed to where you should be able to follow my logic and algorithm. You didn’t give us any tank pressure or temperature, so I have had to assume these values. Your algorithm seems to take an incremental amount of volume resulting in a 5 ^oC storage tank temperature decrease (40 to 35 ^oC). You don’t state the temperature of the buffer tank nitrogen, so I had to assume that as well. The conservative case would be that it is at 40 ^oC; however, I estimated the buffer tank size at 35 ^oC.

You do not state what is the available instantaneous nitrogen flow rate available from your Air Separation Unit (ASU), so I don’t know if the volume of 2.5 m³ is sufficient. You should make the related heat transfer rate calculations in detail and decide what flow rates you need for each tank and the ultimate size of the buffer tank. I recommend that you seek out an available used LPG storage tank for this purpose instead of designing a new, specific vessel. An LPG “bullet” tank should be available, economic (since they are mass-produced), and is usually rated for 20 to 25 barg pressure – which fits in with your available N₂ source of 70 barg.

I would not pipe up the buffer tank N₂ supply as you show it. I would add the N₂ from the buffer tank to each storage tank INDIVIDUALLY AND INDEPENDENTLY. I realize this means more piping and 3 N₂ supply regulators, but these tanks are, in my opinion, critical elements in scope of work and their size alone merits a serious look at the dependability and value of the products they store. I am presuming that you would have a vacuum breaker valve on top of each of the tanks and that this method of breaking a potential vacuum would be reserved for use as a LAST RESORT. I highly recommend that you install a position-detector on the stem or lever of the vacuum breaker valve to generate an alarm signal that is registered AND RECORDED on your control room instrumentation panel. This is done to keep everyone alerted and aware of anytime that atmospheric air has been introduced into the tanks. This, in my opinion, is an important feature because it safeguards the investment and effort put into the quality control and purity of the raw material alcohol and the ultimate product.

I hope this effort and suggestions help you out.

Attached Files

•  N2 Buffer Tank Rev1.xlsx   171.08KB   217 downloads

[DB Shah]

Dear Art

Thanks for the prompt guidance.
I appreciate the amount of the work you input.
I am working on the xls sheet you have prepared and will revert back in case of any query/clarifications.

Regards
Divyang

[DB Shah]

Art-
I went through the worksheet of yours. The tank pressure considered by you is 0.15 Bara (very high vacuum) It should be 100 mmWCg ~ 1.033 bara. If I plug this value it matches my result closely. This being a simple PV relation for ideal gas.


Paul,
You had raised a concern regarding availablity of all N2 in buffer as the PRV pressure will be higher than tank pressure. The thought did flashed me for once but what I think is that at PRV upstream pressure the volume will reduce accordingly. To put it in a simple way-

Point Tank N2 PRV Buffer tank
Pressure ~1 Bara 2.5~ 2.0 Bara 7 Bara
Buffer Vol ~100 M3 100/2 = 50 100/7 = ~14.2 M3

Your second point regarding assumption- yes it has no valid base, just a starting point.
But in next reply I tried to simplify the problem, I can divide my problem in to two parts
1. Rate of N2 reuqired - determined by API 2000, will be useful to size line dia & PRV capacity.
2. Time for which this N2 will be required eg 1 min or 2 mins or more. This will be my basis for N2 buffer tank capacity.

First part of rate of N2 is much simplier. I am struggling for the second part.

Regards
DB Shah