[Occasionally6]

I don't think so but maybe.

As a quick first pass assessment, if you assume:

An air tank volume i.e. V1, of 100 liters (26 US gallons)

Adiabatic expansion (no losses), so not realistic.

Between:

Tank pressure, P1 = 400psi = 2775 kPa

and

Atmospheric pressure, P2 = 14.7 psi = 100kPa.

k for air = 1.4.

Using: P1*V1^k = P2*V2^k

Gives V2 =0.714 m^3

Plug that into:

w = [(P1*V1)-(P2*V2)]/(1-k)

Gives work = 417.5 kJ

At a rate of 4kW = 4kJ/s, that will last a little more than 100 seconds. At 40kW, 10 seconds.