6 replies to this topic

[MattL]

Hi all,

Looking for some advice on a process issue at the plant I work at. We currently have a ~3 gallon tank that is used to "bomb" additives (a solution of optical brightener and 1 gallon of DI water) into our premelt vessels (~800 gallon+) using nitrogen @ 60 psi through a common header. The line ties in about 20 horizontal feet from the tank. The problem is that the tank is on an upper level of the plant and is not convenient for the operators to use/refill. They have to load the tank as many as 8 times a shift, which takes them away from the control room and also exposes them to the elevated temperatures on the upper levels. We are considering relocating the tank to the control room level, about 20 vertical feet below the existing tank. It is piped into a common header to our 4 premelt vessels in 1" sched 80 piping. My worry is that we will leave a significant amount of the additives in the line due to tunneling in the vertical section of the piping. Would adding additional water to the tank improve the transfer? The 20' vertical section should require 0.75 gallons of water to fill. Would a check valve located at the bottom of the vertical riser help? I know it should help with backflow due to material remaining in the line. The transfer itself is automated (valve on tank line, valve at destination vessel, increase in destination vessel pressure indicates charge is complete) but the nitrogen pad on the tank is not (constant 60 psi thru regulator). I would appreciate any help with this issue.

Thanks,

Matt

[Lowflo]

Matt - is this mixture a true solution or a slurry? If it's a slurry (the additives don't completely dissolve in the DI water) then locating the pot at ground level will probably be a problem due to line plugging and poor control over the amount added.

If it's a true solution, then you have a lot of options.
1. Mix the solution in a ground-level tank and use a metering pump for getting the right quantity into the premelt vessel.
2. Same as #1, but transfer the liquid to a charge pot, similar to what you have now, located above the premelt vessel. This will require some modification to the charge pot. To ensure the right amount of solution added to the charge pot each time, you can install a dip tube in the charge pot so it's filled to the same level each time. Once the pot is filled, the dip tube/pipe could route extra flow back to the ground level tank. If the premelt vessel is under pressure, you'll have to pad the charge pot with N2 to make it flow. That adds complexity due to additional controls. That's why #1 is a far better solution - it's simple with a minimal amount of instrumentation.

Regardless of whether it's a solution or a slurry, moving the charge pot to the ground and continuing the current procedures is probably not an acceptable solution. You'll have a hard time getting the same amount of solution in the premelt vessel each time. You just can't maintain a plug-flow of liquid when you're pushing it uphill with gas.

Edited by Lowflo, 25 July 2011 - 04:23 PM.

[Erwin APRIANDI]

Hi Mattl,

I'm still not sure regarding to your system (a simple sketch of current and proposed system would be more easier to understand),
I have quite similar before for processing system of cosmetic product.
For this system I have modified so that the addition of substance is done from under the vessel so I can make a good emulsion.
For transporting the slurry from small vessel I use vacuum pump, so I made a vacuum pressure in the big vessel and introduce the slurry to the vessel by using vacuum effect. (Refer to enclosed file)
I don't know whether this can be applied or not to your system, but hopefully it can add to your ideas

[MattL]

I added a basic schematic of the tank layout. The existing additive pot is at the level of the top of the premelt vessels. The new pot will be 20' below that. The premelt vessels are vented to 5 PSIG or less before the additives go in. The additives are pressurized (from the top of the pot) to 60 PSIG with N2 and flow out of the bottom of the pot, which has a hemispherical bottom head. The pressure in the premelt vessel is monitored for pressure rise to indicate when the transfer is finished. The additives are somewhere between a slurry and a solution due to poor solubility (there is only 10 g of optical brightener in the gallon of water), but we are in discussions with the vendor to see if we can purchase it as a solution or if there is a surfactant we can add to increase solubility.

Matt

 Additive_Pot.xls   47.5KB   44 downloads

[Erwin APRIANDI]

Hi MattL

Thanks for you further explanation, and to answer your question.

Q: My worry is that we will leave a significant amount of the additives in the line due to tunneling in the vertical section of the piping. Would adding additional water to the tank improve the transfer?
A: I'm not sure where you want to put this additional process water, is it means that you will further dilute the additive before sending it to the premelt vessel? If yes, ok the amount of additive left in the piping will be less due to less concentration but still you will have liquid in the piping. Consider to have a drain at the piping low point near additive pot and also flushing connection from the upstream of valve at each premelt vessel if your additive is slurry (to avoid settling and etc)

Q: The 20' vertical section should require 0.75 gallons of water to fill. Would a check valve located at the bottom of the vertical riser help? I know it should help with backflow due to material remaining in the line.
A: Yes it help but only to prevent backflow as you said

Q: The transfer itself is automated (valve on tank line, valve at destination vessel, increase in destination vessel pressure indicates charge is complete) but the nitrogen pad on the tank is not (constant 60 psi thru regulator). I would appreciate any help with this issue.
A: I'm worried if you have this under automatic after you move the additive pot lower than the premelt tank you may have the possibility of plugging (if additive is slurry)
as an example
when the additive fill premelt vessel 1, certain amount of additive will stay in the vertical portion of piping and if this is slurry in certain time the solid will tend to precipitate and cause a plugging problem.
but you can use this sequence for a suggestion.
1. Additive sent to any tank
2. To put a drain at piping near to additive pot location, and to ensure that the additive will all flow by gravity to the drain
3. To use an actuated on/off valve at the drain line, which will open after the tank is fill with additive
4. To repeat the cycle each time of additive filling.

Hope I have explain it clear,
and hope it help

[kkala]

Having read the thread and the useful posts by Lowflo and erwin.apriandi, I wonder whether (α)additive injection into premelt tanks is intermittent (every 1 h ?), (β) further to (control) valves, there is something for even additive distribution into the four premelt tanks. If injection is intermittent, things seem more difficult.
Operators say that existing system has a satisfactory performance, but obliges them to refill the (additive) pot every hour; so this "obligation" had better be eliminated (target).
A probable way out is to make a similar pot close to control room, introduce additives + water there, then transfer the mixture 20 ft upwards by 60 psi N2. Disadvantages have been pointed out in previous posts. Besides steady state in vertical two phase flow is not easy, unless flow patterns clearly indicate mist flow. There is risk/uncertainty at any case (see post by Lowflo, 25 Jul 11).
An alternative would be to "correct" the existing system, by installing an additive preparation pot at ground (near control room) and transferring the required quantity of liquid (after measurement) to the existing higher pot through a metering pump. Capacity of existing pot is 3 gal, solution to be transferred every 1 h is 1.0 - 1.5 gal. A metering pump of (say) 25 gal/h would intermittently transfer 1.5 gal of solution in less than 4 min through a transfer line of 1/8 in (*). This transfer line (connecting the two pots) will be full always; but if the solution contains suspended solids or the composition of optical brightener can be (partially) destroyed, transfer line (of negligible volume now) can be purged to ground pot (manually or automatically), after any operation of metering pump.
The alternative is represented in the attached "Additive_Pot_alt.xls". It is similar to proposal No 2 by Lowflo (25-7-11). Of course details of the installation not known (or not "detected") may render this alternative infeasible.

(*) In refineries (and other industries) lines of dia < 1" are not used. In this case transfer line should have minimum diameter to limit liquid content in it. Special supports may be required.

Attached Files

•  Additive_Pot_alt.xls   46KB   30 downloads

Edited by kkala, 31 July 2011 - 11:31 AM.

[kkala]

Following previous post of 31 Jul 2011, new additive preparation pot (on ground) had better be an agitated tank to prevent settling. It has been understood that there are suspended solids (undiluted additives) in the solution, even though solutions from their supplier may be used in the future. At any case agitation would homogenize the liquid mixture, too. So attached "Additive_pot_alt.xls" is slightly modified to show this agitated tank. As noted before, details not known to me could lead to another solution.

Attached Files

•  Additive_Pot_alt.xls   45KB   23 downloads