Oil Skimmer tank design - Nutshell filter Design

[mattegm]

Goodmorning,
i have to design a oil skimmer tank upstream an hydrocyclone package.

I have 100 m3/hr of process water containing 1000 ppm oil (max) that enters an oil skimmer tank in order to eliminate in first stage a part of the oil that is contained in the water ( just to reduce if possible, the quantity of oil going to hydrocyclones).

I have a preliminary P&ID (attached) in wich is clear that:
- the water/oil enters the tank (P=2 barg)
- the oil on the surface of the water is collected in a "oil skim chamber" and directed to nutshell filters
- The "partially deoiled water" is directed to hydrocyclones with a booster pump

Can anybody help me to design this kind of item?First of all, i think i can estimate a residence time (i.e. 20 minutes) but i don't know how to design the internal items.

[link Point to another website Only the registered members can access]
Uploaded with

[link Point to another website Only the registered members can access]
Thanks

[mattegm]

No ideas?

[mobek]

mattegm:
These things require intellect, time & energy to do. I can certainly help you BUT for FEE. If you agree then we've to sign a contract.
Do you agree?

Best wishes.

[mattegm]

mattegm:
These things require intellect, time & energy to do. I can certainly help you BUT for FEE. If you agree then we've to sign a contract.
Do you agree?

Best wishes.

I mobek, thanks for your reply and interest.
Do you work for a company that supply this kind of items?
In that case you can tell me the name and we can do something "official"..

Regards

[mobek]

I'm an independant consultant. I did a number of these designs and much more sophosticated process vessels design.

[TOLGAUS]

If I were you I would obtain my oil particle size distribution curve first. This will dictate all restriction & constraints in your design. if your feed does not change with respect to water chemistry, then you an proceed. Otherwise, you need to provide maximum residence time possible with calming baffles. I really recommend calming buffles (OPUS maxim is expert in this). Then i would recommend internal inlet flow devices. The weir height is a very important concern in these devices; keep it as high as possible. Move the weir as far from the inlet as possible. use vortex breakers for the water outlet. Finally verify your design with CFD analysis. Then you are done almost.

[TOLGAUS]

if you can provide an efficient design in this skimmer, I guarantee you that you can bypass the nutshell shit and have some cost reduction in maintenance. Nutshell filters are some huge and useless most of the time.

[mattegm]

If I were you I would obtain my oil particle size distribution curve first. This will dictate all restriction & constraints in your design. if your feed does not change with respect to water chemistry, then you an proceed. Otherwise, you need to provide maximum residence time possible with calming baffles. I really recommend calming buffles (OPUS maxim is expert in this). Then i would recommend internal inlet flow devices. The weir height is a very important concern in these devices; keep it as high as possible. Move the weir as far from the inlet as possible. use vortex breakers for the water outlet. Finally verify your design with CFD analysis. Then you are done almost.

Hi tolgaus,
first of all: the hydrocyclone downstream the oil skimmer are designed to reduce from 1000 ppm to 25 ppm of oil. The 1000 ppm is also the inlet basis of design of the oil skimmer so, at the end, the oil skimmer is not really necessary. But the client want it as ""safety issue" so we have to provide them.
For this reason we decided, with the client, to design the oil skimmer only with a sufficient time of residence.

I agree with you to use calming baffles and the maximum residence time possibile. Do you think a vessel of diameter 2400mm and 10000mm long with sufficient baffles will be ok to do this work? I remind that the purpose of the oil skimmer is only for safety, the downstream equipment can foresee the specification also without the oil skimmer.

Thank you tolgaus, if you agree i will ask you more as soon as i proceed qith the basic design..
thanks

[TOLGAUS]

Hi mattegm,
I agree with you that normally hydrocyclones can receive 1000 ppm and produce an underflow with 25-30 ppm if the condition* below is satisfied:
*the optimum flow per liner and a reject/underflow dP ratio between 1,6 - 2,5 depending on the type of the liners. I presume that the booster pump and the cyclones are already designed according to these parameters.

I also presume that since you are mentioning the 'safety issue', either you have a spec brake across your feed to the cyclones and actuallly you want to use the oil skimmer like a knock-out drum provided with a PSV (as per P&ID) OR your client wants some extra residence time to ensure that the required OIW quality is always achievable. Please clarify this point.

Moreover, unless you are dealing with an oil with very low API, it seems 20 minutes residence time are more than enough. I normally treat my produced water (inlet 30 ppm - outlet 5-8 ppm oil density ~750 kg/m3) with a residence time of 3 - 4 minutes (water clarifier and demulsifier in injected upstream) with a simple oil skimmmer (degasser). 20 minutes would be a gift

For your information, to my knowledge, hydrocyclones are only capable of removing oil with diameters larger than 10 microns. So, if I were you, I would certainly check the expected particle size distribution of my produced water feed just to be sure before kick-off the design work.

Furthermore, it would be a good idea to reduce set pressure of the skimmer as low as possible to the limit where it corresponds to the maximum NPSH required of your booster pump. The more degassing you have, the more natural gas flotation you will apply.

Based on your design constraints, I understans that your skimmer is actually 25% loaded as per the desing flows you specified. So, once you install the calming baffle plates, a simple inlet flow distributor, a V-notch type skimmer, carefully calibrated level instruments and a small dosage of demulsifier injection upstream the skimmer, your skimmer shall work quite OK. Note that you may even not need any internals I specifed since 10 meters is alredy a very good distance to calm the water. But, this depends on the budget you have and the conditions of your contract with the client. Also to the water chemistry. Is there any emulsion expected in the skimmer?

Also, have a careful look on where the reject stream of the cyclones are routed normally. Sometimes they can be routed to the upstream of the feed and you have to consider this negative impact in your design.

Finally, what is the fuction the nutshell filters? Normally I always considered them as polishing units but not a unit to treat the skimmed oil from the oil skimmer. Maybe I will learn something.

[mattegm]

Hi mattegm,
I agree with you that normally hydrocyclones can receive 1000 ppm and produce an underflow with 25-30 ppm if the condition* below is satisfied:
*the optimum flow per liner and a reject/underflow dP ratio between 1,6 - 2,5 depending on the type of the liners. I presume that the booster pump and the cyclones are already designed according to these parameters.

our supplier of hydrocyclones designed them considering the process specification, i hope they considered all these conditions

I also presume that since you are mentioning the 'safety issue', either you have a spec brake across your feed to the cyclones and actuallly you want to use the oil skimmer like a knock-out drum provided with a PSV (as per P&ID) OR your client wants some extra residence time to ensure that the required OIW quality is always achievable. Please clarify this point.

you are correct al 100%. This is the purpose of the oil skimmer.

Moreover, unless you are dealing with an oil with very low API, it seems 20 minutes residence time are more than enough. I normally treat my produced water (inlet 30 ppm - outlet 5-8 ppm oil density ~750 kg/m3) with a residence time of 3 - 4 minutes (water clarifier and demulsifier in injected upstream) with a simple oil skimmmer (degasser). 20 minutes would be a gift

Ok, if you say this i can size the oil skimmer with a lower time of residence, so that i can reduce the size.
Now, just considering a residence time of 11 minutes in the section before the "skimmin baffle" and considering a filling of 70% of liquid i have a tank of ID 2400xTL/TL 8000.
I divided the tank in the section of "oil skimming" (residence time 11 min) and the section of "water outlet chammber" (residence time 5 min).
DO you think is correct?Or i can reduce more the size?

For your information, to my knowledge, hydrocyclones are only capable of removing oil with diameters larger than 10 microns. So, if I were you, I would certainly check the expected particle size distribution of my produced water feed just to be sure before kick-off the design work.

In my specifications, the outlet of hydrocyclones is <10 micron, so i think before the hydrocyclones it is >10 micron...I hope!

Furthermore, it would be a good idea to reduce set pressure of the skimmer as low as possible to the limit where it corresponds to the maximum NPSH required of your booster pump. The more degassing you have, the more natural gas flotation you will apply.

Atmosferic would be ok?

Based on your design constraints, I understans that your skimmer is actually 25% loaded as per the desing flows you specified. So, once you install the calming baffle plates, a simple inlet flow distributor, a V-notch type skimmer, carefully calibrated level instruments and a small dosage of demulsifier injection upstream the skimmer, your skimmer shall work quite OK.

ok, good.

Note that you may even not need any internals I specifed since 10 meters is alredy a very good distance to calm the water. But, this depends on the budget you have and the conditions of your contract with the client. Also to the water chemistry. Is there any emulsion expected in the skimmer?

The budget is low, considering the fact that they "stole" this equipment to us))
I forgot to tell that upstream the oil skimmer there is a dosage of reverse demulsifier and corrosion inhibitor.

Also, have a careful look on where the reject stream of the cyclones are routed normally. Sometimes they can be routed to the upstream of the feed and you have to consider this negative impact in your design.

the reject stream of hydrocyclones is directed to closed drain header, so i think it is not a problem

Finally, what is the fuction the nutshell filters? Normally I always considered them as polishing units but not a unit to treat the skimmed oil from the oil skimmer. Maybe I will learn something.

In this project, nutshell are downstream hydrocyclones in order to reduce oil concentration from 25ppm to 10 ppm. I foun this "approach" in some other project for offers that i've seen...

Thnak you so much for your help.

[Dr.Process]

Dear mattegm & TOLGAUS,

Do you guys know how to size a hydrocyclone downstream of production separator? Produced water will be used for water injection

Thanks

[poojathakurhmr]

Mariners Planet is job portal site for all types of jobs on ships, shipping jobs, marine jobs and sea jobs for Seafarers, Sailors, Mariners, Shipping Companies and Marine Institutes.