Elmer for Oil Spill "Solution" Assessment

[arit]

Hi guys,

[for the hastly reader the important parts are highlighted]
I am sure you have heared about the oil spill in the gulf of Mexico and the inability to stop it. You also might have heared that they are looking for suggestions (http://www.deepwaterhorizonresponse.com ... 931/541843).
I have been thinking about trying FEM software for a long time but never had a chance to. Now I thought that a crackpot idea in regards to the oil spill could provide the excuse to actually do a hands-on "experiment". So here is is the plan for stopping the leak:
Inserting an array of needles into the leak and cool them down by thermoelectric cooling. The seawater and the oil freeze around the needle, lowering the leak rate. The frozen layer grows thicker, eventually stopping the leak.

Now independently of wheter or not this is a possibility to actually fix the leak I think the assesment of feasability provides an interesting enough multi-physics problem. You have the needle array, the flow out of the leak, the phase change and different materials.

So here is my question: Is this a problem that could be tackled with elmer? How much time would you expect a newbie to require to get an answer to this problem if he is willing to simplify this problem a lot (2d or even 1d geometry etc.)?

Thanks and sorry for the long (and slightly crazy) posting.

Cheers,

arit

[raback]

Hi arit

This case seems to be still actual so I'll give my 2c for commenting your case: I would not rush into FE analysis. Some rough manual estimates might be enough to decide on the feasibility of the idea. In 0D analysis you could estimate the heat capacity of the leaking oil. Then knowing the efficiency of the cooling process you might estimate how much electric power should be available at the site.

If the idea still seems good you could try to look at the problem at the scale of the needle. The needle should be pretty long to give enough range between the hot and cold ends. Then there might be that the heat transfer at the boundaries would totally dominate the temperature distribution on the needle. This is something that could be solved with FE analysis. However, Elmer does not directly support thermoelectric simulations but they could probably be accounted for with the user defined subroutines...

-Peter

[arit]

Hi Peter,

I finally did the order of magnitude estimation. It turns out that less than about 10 kWh/(kg of oil) are required to electrically freeze the oil. This means that 3% of the power yield of an aircraft carrier would suffice to freeze 1/1000 of the oil leaking out right away. As the goal is to cloak the leak there is no rate requirement, so it sounds possible.

So if I want to continue my travel here, how would I start getting at the problem at the scale of the needle?

Cheers,

arit