It's a normal speed 9-man (I see there are 13500 total chips in play) and the blinds are already 150/300*, aren't they? (The previous blind levels are 50/100 and 100/200 on iPoker, and a 4.5x steal would look weird from a reg.)
In general, I use Holdemresources Calculator for this. There's an option of 'advanced analysis' of Nash resteal ranges there, as a first approximation (to maximise the ROI, one still needs to adjust to opponents' styles and exploit them, which can be analysed too, by editing opps' ranges manually).
You've inspired me to run a simulation there after a long hiatus
* Edit: only judging by the amount to which the converter says Hero raised, I've figured out that the blinds are 150/300 and not 100/200, and the converter is way too buggy (doesn't specify blind sizes and uses the 'raise by' mode calling it 'raise to'). Please post hands using PT/HM (copy to forum format) instead - there are instructions in the 'important' thread 'How to use hand eval forums'.
The below spoiler was written for the case of 100/200 blinds, still let it remain here (read it for an intuitive explanation):
Spoiler
Then HRC shows that, in a vacuum (and SB does look like an average TAG reg knowing that ICM charges a 'tax' on equities in flips but is also the reason why his minraise has a bigger fold equity), we should shove wide - 44.4% (SB is afraid of becoming short as much as we are of busting out):


(the numbers are the percentages of the prize pool gained or lost as opposed to the next best option, i.e. folding in our case.)
The thing is that, with these stack sizes (22 bb effective), people shove low pocket pairs, aces/kings with weak kickers, (semi)connectors - they do reasonably well in all-ins, but not stellarly, most likely flip 55/45 or 60/40 vs medium card hands, and so they won't fold to a resteal but need to gain additional fold equity from hands that would resteal and be in a good shape.
On the other hand, they have to play a super wide range because of the chip lead, and the still strong ICM effect with the 50/30/20 payouts, so they should raise/fold a lot here (and raise/call biggish aces, premium kings and biggish pocket pairs).
So KQo is too good to fold - a big chunk of hands that it does meh against (Ax, PPs, vs which you wouldn't wanna flip because of the ICM) is excluded from the minraising range because it's openshoved instead; vs most of the remaining hands, KQo does very well just because they fold, and vs those rare hands that raise/call, it's still in a decent shape.
Vs a typical iPoker reg, though, we need to resteal narrower than Nash (his range is stronger and we're risking the whole stack to grab his 2.25 bb and our 1 bb). But KQo is still a shove. In the case of a nitty SB who doesn't steal BvB remotely often enough:

In the case of 150/300, the conclusions are roughly the same:
Spoiler


Vs a nit:

Bottom line: it's hard to fold KQo this deep!
Or did you mean you could use the position power and flat? Well, you could certainly do so if you had reads on the opponent, like his habit to double-barrel AI too seldom and give up to float bets too often. But usually with 15 bb effective the position doesn't matter that much, and we end up missing the board and folding most of our preflop flatting range which (unlike specifically KQ) seldom has even two overcards that would justify floating...
If someone enlightens me how to play this postflop after flatting, I'll eagerly learn