Since I design a lot of repairs on wastewater treatment tanks, by all means use the 1.4F and try to optimize the test case. We can always use the business.
Or you can do what we do and use 2.21 times the fluid loads for bending and 1.7 for shear to whatever you have as a passive overflow, like the tops of walls.

Thanks Jed. But could you guide me where how to get 2.21 for bending? is this about durability factor? Sorry im not good with tank design.

Also, as with this Sd, i read from ACI350 that Sd could be ignored when load combination is with seismic. I was thinking maybe the reason was that seismic load is only short term. If my intuition is correct, then i could also ignore Sd for wind and for test load, since they are not long term forces. I may be wrong, would be thankful to be enlightened

First of all, ACI 318 is wrong. Fluid loads, while potentially well defined, should have a much higher load factor than dead loads. ACI 350, some edition, somewhere, says to use a 1.7 factor and add a 1.3 factor to that for "durability" for bending. Shear requires a 1.7 factor, unless you need shear reinforcing, than you use the 2.21.
You're right, S_D can be ignored when combined with seismic. But seismic is likely to not control vs. the overflow level on the vessel.
One last piece of advice; these tanks, full of wastewater or water, are not to be trifled with. I can't mentor you, but someone should be watching.

Thanks a lot for this guidance Jed

I think I'm following your logic, but please check me: Since these are test cases (not long-term operations), do you ignore seismic added stresses for the four-five hours of the static test for waste tanks? (I'm more familiar with the pipe and ASME type hydro tests, op tests.)

Leak tests last a lot longer than four to five hours, but the principle is basically what you stated. Based on the risk of a seismic event occurring during the leak test, it can be reasonably ignored.