10 replies to this topic

[mortino]

In a utility center of an IGCC power plant,we want to do N2 blanketing over two demin water storage tanks each 2120 m3 with design pressure of -6/+20 mbarg,maximum pumout flow rate is 300 m3/h,I assumed to use bottled nitrogen for first year of the plant operation and later on an air separation unit will supply our consumption.

Does any one know a reliable technical source to do the best estimation on how many bottles or volume of liquid nitrogen we need?

Highly appreciate inadvance

[Art Montemayor]

Mortino:

The way I've always estimated N₂ consumption - and also the way I've known others do it also - is by empirically calculating the volume of tank liquid displaced.

In other words, since all a blanketing system does is maintain a "blanket" over the stored liquid, the theoretical "consumption" of N₂ is ZERO. However, we all know this is not true. It isn't true simply because there is a need to replace the N₂ that is displaced and forced out of the tank when the tank level is increased (after being decreased). And the only reason a tank's level is varying is because of the "dead band" of liquid level that has to be maintained on the tank's level control due to the need for control flexibility and instrumentation reaction. If it were a perfect world, you could set a tank's liquid level at any height and design a feed pump to exactly feed the flow rate into the tank that exactly matches the exiting flowrate out of the tank. This is practically impossible and consequently what you probably have is a tank level control. When the tank drops in level due to water consumption, a water feed valve opens and allows makeup water to enter the tank after the level reaches a set low level height. The makeup water continues to fill the tank and when it reaches a set liquid level, the makeup water is turned off.

When the tank water level is decreased, the tank's N₂ blanket pressure is decreased (since it's volume is increased) and a possible partial vacuum is possible unless makeup N₂ is fed into the tank's vapor space to maintain the blanket pressure at a set point. Now, when the tank's water level is increased subsequently (as explained above) the N₂ blanket is compressed and its pressure is increased (since its volume is decreased). This cannot be allowed to happen beyond a certain fixed pressure that is related to the tank's maximum allowable working pressure (MAWP). Therefore, the previously fed makeup N₂ has to either be recoved by "sucking" it out of the tank with a compressor and storing it in a compressed state - or by simply venting this excess N₂ to the atmosphere. The latter is usually the accepted and followed case since most times the amount of N₂ vented does not justify a return on the investment for a compressor and gas storage facilities.

Knowing the above, therefore, the key to calculating an estimated N₂ venting quantity (or "consumption") is to estimate the number of cycles the tank goes through in changing its water level. This you can estimate empirically or by estimating the cycles of water demand and/or its frequency. The actual calculation of the N₂ volume is easy since you know the tank dimensions. All you need to estimate is the number of level change cycles in a given day or other convenient time period.

If you fine-tune your process water requirements and keep your water level dead band narrow, you should be able to minimize your N₂ venting or "consumption".

Of course, all the above is assuming that there are no N2 leaks in the tank - which is not 100% true. I usually add a 5% contingency to allow for leaks.

[proinwv]

Allow me to add to Art's comments. The consumption due to leakage can be more than one might anticipate. Many tanks are simply not tight, regardless of what some users might like to think or say. The leakage from these openings is continuous 24/7 and can be significant. One way to determine if the tank is tight is to monitor the N2 flow. It should be zero except during tank cycles. Another way is to use a blanketing valve with instruments that indicate whether the valve is open.

Also, there will be some consumption due to cooling.

Last, I have never heard of an operation that was pleased with bottle gas because they underestimated usage and the costs were much higher.

[Art Montemayor]

Mortino & Others:

Paul has brought out some very important points that I feel I failed to mention or give importance to and I want to make up for my flawed post:

1. Leaks related to storage tanks can indeed be significant. Even the best of conservation vents and other relief devices leak. They are certainly not "bubble-tight" - nor do they claim to be. The usual conservation vent is actuated by a gravity-weighed pallet that many times does not seat perfectly well and allows vapors to escape. What may seem to be an insignificant leak rate could add up to be well in excess of 500 - 1,000 scf of N₂ in one day's operation.

2. The above real-life possibility is one reason why employing compressed gas cylinders (with a capacity of approximately 300 scf N₂) doesn't result in a very practical or economical application when one considers the quantity of cylinders and the amount of replacements that have to be made available. As Paul correctly points out, H.P. cylinder N₂ is just not a cost-effective solution as a source of N₂.

3. With the ever-increasing costs of N₂ (due to increasing energy costs) the need to monitor and reduce N₂ leaks and losses continues to play a big role in keeping total costs under control. The economic generation of gaseous N₂ is highly dependent on energy and power costs.

4. The cooling of storage tanks that usually causes "in-breathing" in atmospheric-vented tanks causes a pressure reduction in N₂ blanketed tanks and thereby triggers the addition of make-up N₂ to maintain a set tank vapor pressure. Upon subsequent heatup (during the daylight hours) the N₂ expands, increasing the vapor pressure and triggering a vent of N₂ from the tank.

[Nirav]

Does any one know a reliable technical source to do the best estimation on how many bottles or volume of liquid nitrogen we need?

hi mortino,

You must refer to "in breathing" and "out breathing" requirements of tanks as provided in API-2000.

It provides very useful guidelines in detail to estimate these requirements.

I don't think using N2 bottles would be a good idea. The reason is already explained in earlier responces to your query.

Thank you,

[mortino]

Dears Art, Paul and Nirav,

Thank you for your comments.

Nirav,don't you think that API 2000 is a very coservative recommendation?

Regards from Morteza and Happy new year

[proinwv]

API 2000 is very conservative. Read the Appendices for their rationalization. Part of it is that vents are very cheap as compared to tanks, and their contents.

I do know that some users reduce the API values to their "in-house standard" but that should be done with some factual basis and caution.

Best of 2007 to all.

[jprocess]

Dear Mr.Montemayor,
You stated that :
"...the key to calculating an estimated N2 venting quantity (or "consumption") is to estimate the number of cycles the tank goes through in changing its water level.
This you can estimate empirically or by estimating the cycles of water demand and/or its frequency.
The actual calculation of the N2 volume is easy since you know the tank dimensions.
All you need to estimate is the number of level change cycles in a given day or other convenient time period."
Now my question is that how can we estimate the number of cycles?I have not any sense about your statement.
Also as you know API 2000 and Ludwig chapter 7 have tables that relate the rate of required blanket gas to vessel volume.Are these values exact?Can we use them instead of using your method?
Thanks in advance.
Warm Regards.

[Technocrat]

By the way, what is necessity of blanketing the DM water tank?

[proinwv]

In reading API 2000 you must, in this case, recognize that this is a benign, product, not a fuel nor a liquid with a high vapor pressure.

As previously stated, the product exchange rate of the tank is the paramount issue for sizing. Vaporization during filling should be nil.

Next consider the effects of sudden cooling, especially due to sudden heavy rain, and also some overnight cooling. You certainly won't require the full API requirements which are for a high vapor pressure product.

Good mechanical tightness, adequate supply of N2, and proper sizing are the key elements for this application.

[fallah]

By the way, what is necessity of blanketing the DM water tank?

Dhirajkumar
In steam power plants DM water is used as make-up water, and because the boiler water is subjected to high P and T if the make-up water absorbs oxygen from the atmosphere, then the conditions have been provided for higher corrosion rate.Of course, any dissolved oxygen is continuously removed from water cycle physically by steam injection (in Deaerator ) and chemically by injection of an oxygen scavenger(such as Hydrazine).
Fallah