Comments on your answer:
A) The math is correct, but the units are not. When you multiply gallons per hour by hours, you get gallons. The result should be expressed as 29.6 Mgal (per 24-hour period).
B) The first and second parts of the curve are concave with respect to your estimating trapezoid, so your estimate can be high in those time periods. The third part of the curve is convex relative to your trapezoid, so you under-estimate that part. Overall, it looks like your approximation is an over-estimate.
C) Yes, 1.233 Mgal/hour (don't forget the units)
D) The curve has a change of slope at or near t=12. The slope at t<12 is negative, while at t>12 it is positive. If this were an absolute value function, we would say the slope is undefined at that point. If you don't simply want to declare the slope undefined, you can use an estimate from 10 to 14, which will give a slightly negative result: (1.4 -1.5)/(14 -10) = -0.025.