12 replies to this topic

[daryon]

Hi All,
I've just been involved with commissioning a seawater system on a FPSO. The SW water lift pumps are located in the ships engine room a meter or so above the keel. The pumps take suction from the sea chest and pump to the process modules which are 3.5 m above the upper deck, and a total of appox.25 m above the pumps. The main consumer for SW are the two vacuum condensers which use 1300 m3/hr each.

There is a 26 inch SW line which dumps SW overboard once it has passed through the condensers. On this overboard dropline there is a 6 inch vent line which I guess was installed as a vacuum breaker. It's not working. The dropline runs down the side of the ship 10m and total length is about 15 m, the falling water is resulting in a partial vacuum at the outlet of the condenser (-0.4 barg). You can hear the 6 inch vent whistling away as it sucks in air with quite some velocity. This vent is definitely undersized, but what i'm trying to work out is by how much. I'm thinking to suggest they increase the vent size as when the flow is at 2600 m3/h the vibrations are noticeable throughout the module. The STG vendor commented that is is not good for the turbine or generator.

For self venting flow the drop line should be 42 inch, but it can't be replaced easily. A simpler modification is to install a suitably sized vent. Can anyone give me an idea how to size this vent? Reading through a previous thread someone suggested to
- 1st guess the required air flowrate then calculate the 2-phase pressure loss in the dropline
- Calculate the static head pressure in the dropline based on the 2-phase density
- Adjust the air flowrate until the static head pressure and 2-phase pressure loss along the pipe are equal
- Size the vent to ensure it can supply the required air flow at a reasonable velocity.

DO you think this method works? Does anyone have any other procedure for sizing the vent?
Appreciate your thoughts
thanks

[shan]

I think you should close the venting valve to prevent air into the system.

[Zauberberg]

It's difficult to follow your presentation without a clear sketch of the system. If you don't mind, please take a couple of minutes of your time and I'm sure we can understand better what is going on there and provide you with some comments.

[daryon]

Here's a rough sketch, please forgive the presentation I wish I had a little more time to present it neatly. Any questions let me know. I'm really just looking for a procedure to size a siphon/vacuum beaker line.

Attached Files

•  SW Cooling System Sketch.pdf   274.06KB   410 downloads

[Zauberberg]

Obviously there is an error in design of the drop pipe: 15m static head creates excessive vacuum which cannot be accommodated by the given vent line size. You have the following options:

1. Cut the drop line if you can allow for the sewater discharge from bigger height.
2. Provide a manual backpressure valve as close as feasible to the bottom end of the drop pipe (eliminate the vacuum), and throttle it to achieve required pump performance. You will likely damage butterfly valves at the outlets of vacuum condensers if you keep them 10-20% open continuously.
3. As per your sketch, you need additional air inflow to compensate for approx. -0.20 barg vacuum (-0.40 barg plus the 2m static head). Normal practice is to have vent line at least one half size of the drain line or bigger, which translates into 12" vent line size.

[Zauberberg]

Another elegant solution would be to build a small pond somewhere in the horizontal portion of the drop pipe (simply cut the drop pipe, discharge it into an open pond, and then connect the vertical portion of the drop pipe to the liquid outlet from the pond). That way, vacuum could never be created.

I think that closing the vent will not prevent creation of vacuum since there is no backpressure device (orifice, valve) at the outlet of the drop pipe. As long as there is 15m vertical pipe discharging into atmosphere, negative pressure will occur on the top of the pipe.

The best solution depends on the amount of space and time you have there. Installing a manual globe valve coupled with orifice is perhaps the simplest one, installing a pond is the most elegant one. There is an option to cut the vertical section of drop pipe as well. However, you may opt for the bigger vent line size if the first two solutions are not feasible.

Edited by Zauberberg, 08 June 2010 - 05:41 AM.

[shan]

In my opinion, the purpose of vacuum elimination is to create backpressure to allow the pump on curve at 2600 M3/Hr. If my assumption is right, you have the following two options.
1. Add a by-pass line to dump the excess water back to sea at the pump discharge. The energy waste is not much if your pump curve is not that flat.
2. Install a variable speed driver on the pump to achieve the desired rate at whatever backpressure.

[katmar]

I suspect that daryon is quoting a method originally proposed by me to calculate the quantity of air drawn in and for sizing the vent. I still think that method is correct if the purpose is to stop the vent making a noise, but as long as there is two phase flow in the final downleg there is a possibility of vibration and this should be avoided.

There are two ways to prevent air being drawn in. The first is to simply close off the vent, and the second is to make the downleg self venting.

If the vent is simply closed off and no other changes are made there will still be vibration because the water will draw a vacuum and there will be boiling (vaporization) and cavitation (re-condensation) as the water travels down the leg. This is why Zauberberg (in his second option) has proposed providing a back-pressure valve at the final outlet. The only disadvantage I see with this method is that if the turbines and their condensers are sometimes run singly the outlet valve would have to be adjusted according to the flow.

The second method to avoid air being drawn in (i.e by making the downleg self venting) would require either a massive pipe for the entire downleg (I agree with daryon's calculation of 42") or a self venting section of pipe at the top of the downleg with a control valve at the bottom of the leg to maintain the liquid level in the self venting section. This option is probably more complicated than is warranted in the situation.

The "small pond" option suggested in Zauberberg's second post is in fact what was done routinely in process plants in the olden days. An open channel was provided to collect the discharge from all the condensers and then this would be led to a pipe that led back to the cooling tower. In this scenario it is still possible to have air drawn into the final downleg and there could be vibration in that pipe, but the vibration would not be hydraulically transferred to the condenser outlets. If the down-pipe was properly supported the vibration would not be bad. I suppose an open channel is no different to a massive vent pipe.

[shan]

If your guys persist on introducing air in, I think to it is better off to connect utility air line or to install an air blower than to increase the vent line. If you double the air pressure, the flow velocity will be equivalent to double the pipe area in the condition of the identical mass flow. Right?

[daryon]

Dear all,

Thanks for the replies, it's great to get your opinions and thoughts. Katmar you're right it was one of your posts that suggested the method I mentioned in my initial post, and to me it seems logical. Using that method i'm estimating at least a 16 inch vent line, based on sucking in 17,000 m3/h of air and keeping the air velocity in the vent line below 40 m/s.

Zauber, I like your elegant 'pond' solution but unfortunately no scope to implement this design on a floating production unit. The best we can do is properly sized self venting line or a adequate vent. I'm interested in the basis for the rule of thumb you mentioned "Normal practice is to have vent line at least one half size of the drain line or bigger"

Another solution we are mulling over (not too dissimilar from the 'pond' idea) is to cut the drop line and install the collection funnel, in addition to increasing the vent size. (See attached schematic), this seems to be a good way of isolating the pipe vibrations from the process module. When 2600 m3/h is falling down the drop line the vibrations are noticeable all over the process module, this is not good for the steam turbine especially the coupling with the generator. Unfortunately we can not just cut the drop line as the vessel weather vanes which means the prevailing winds will always blow the seawater back at the process, and its a lot of aerated seawater to be spraying over expensive equipment.

Also I think installation of an RO should be considered to add resistance to the system, i'm concerned for the life of the butterfly valves at the vacuum condenser outlets if they are maintained at just 10 to 20% open. Either the valves will erode or the pipework immediately downstream of them will. Unfortunately a VSD for the pump motor (to get the required flow at a lower head as per Shans suggestion) is much too expensive.

Always interested to hear your thoughts, thanks again

Attached Files

•  SW Overboard dump sketch.pdf   21KB   238 downloads

[Zauberberg]

I like your sketch and particularly blue ocean waves around the FPSO. It gives the complete feeling.

Yes, funnel or pond will work - essentially it's the same concept. In addition to that, I would certainly opt for a back-pressure device at the very end of the drop pipe (just prior to discharge to the pond). Globe valve would be ideal, I guess. It gives full control ability in cases when one or both condensers/turbines are working. And it will protect the butterfly valves downstream of vacuum condensers.

Increasing vent line size should be re-evaluated for this concept because it could save time and money. Depending on the new height of the drop pipe (upstream of the pond), the quantity of air required will be proportionally less than it was the case originally. Perhaps you can go with the existing 6" size.

Good luck,

[chemsac2]

Zauberberg,

Spot on!

I came to same conclusion yesterday when I read Daryon's lastest comment. But then thought otherwise.

My argument is:

Driving force would reduce proportional to drop height (Static head proportional to h).

However, frictional pressure drop would reduce proportional to height as well (DP proportional to h). Any other parameter in pressure drop equation (DP = 2*f*L*u2*rho/D), does not change except may be friction factor a bit.

Thus, air entrained would not be any different from that for 15 m drop height.

Can you back your statement with an explaination? That would be very helpful.

Regards,

Sachin

[Zauberberg]

The explanation is quite simple: imagine if we reduce the drop pipe height to 0m. Do we need any vacuum breaker then? Definitely not.

Static head of water or air+water is what creates vacuum in an enclosed system, when discharging to atmosphere. If we eliminate the static head, we are eliminating vacuum as well.