Now I am not getting it physically why should there be any Coulomb energy in this case as there is no another charge to provide the electrostatic coulomb potential energy?
You're right. For point charge the formula
$$
\int \frac{1}{2}\epsilon_0 E^2 d^3\mathbf x
$$
gives infinite value and is thus unusable - infinite energy would mean one cannot do calculations with it. Indeed, one can derive the above formula only for regular (finite density) distributions of charge, not for point charges. For set of point charges at rest, based on Coulomb experiments we define energy as
$$
W = \sum_{a}\sum_b' \frac{1}{4\pi\epsilon_0}\frac{q_a q_b}{|\mathbf r_a - \mathbf r_b|}
$$
(primes means only those $b$ are used for which $b\neq a$). This formula can be transformed into
$$
W = \int \sum_{a}\sum_{b}'\mathbf E_a \cdot \mathbf E_b \,d^3\mathbf x
$$
where $\mathbf E_a$ is field due to point charge $a$, and so on. This formula can be generalized to case when charges are moving (analogous magnetic term is added) to obey both Maxwell's equations and the Lorentz formula for EM force. See the paper
J. Frenkel, Zur Elektrodynamik punktfoermiger Elektronen, Zeits. f. Phys., 32, (1925), p. 518-534.
http://dx.doi.org/10.1007/BF01331692
If the charge is alone in large region, according to this formulae it has zero EM energy since the summation contains no term; its only energy is the rest energy $mc^2$ due to its mass.