Hi.. :)
"Do bats eats cats" has correctly pointed out that.
Or you can think of the question as:
Lottery 1: Giving 2,50,000 with probability "p" and 1,50,000 with probability "1-p"
Lottery 2: Giving 2,00,000 with probability 1.
Now, given expected utility maximiser you'll prefer Lottery1 over lottery 2 only when Expected utility from lottery1 is greater than expected utility from lottery2.
ie. p*log(2,50,000)+(1-p)*log(1,50,000) > log(2,00,000)
:)