You can install nitrogen as a blanket.  Your concern should be to ensure relief is sufficient in the event of, for example, the pressure regulator failing and mains nitrogen being applied to the tank.  The pressure of the blanket should not be any different to that achieved during inbreathing and outbreathing during emptying and filling.

As long as you have sized and installed your relief system satisfactorilly and routinely examined it then all should be well.

Adam Potter MEng CEng MIMechE
www.ax-ea.co.uk

for this i will use API2000,

but if you have a tank without anchor bolt, you are limited in the pressure (API 650 AP F).

Does anyone have a opinion about this?

We design API 650 tanks all the time with N2 gas blankets. An example would be a 100,000 BBL tank, design pressure 3.5 kPa, tank connected to a VRU, gas blanket pressure kept below PVRV set pressure which for this case woulod be approx. 3.0 kPa.

If your tank is not anchored and you have an internal pressure then your shell might be subject to uplifting with or without additional wind uplift loading.   You must do the calcualtions to check this otherwise you will damage the bottom shell to floor weld and then you're in big trouble.   Appendix F checks your roof to shell joint for failure and that also has to be check as well before you check the design of the existing tank to have an internal gas pressure.

A nitrogen pad is only one of many ways to generate internal tank pressure.
It seems to me the tank anchor issue is independent of nitrogen padding.
Do you feel a nitrogen pad will change the tank's rated pressure?  

but API 650 is limeted to 17,2 Kpa
in adition with Apendix F point F1.2

if the pressure is hiher we should use API 622