more changes to FTA. Need to fix tests.

Changed convolution calculation in __risk function to multiply with haz_freq_deltas instead of hazard_freq. Final risk makes a lot more sense now, but still different from CCY results from NED paper. Need to perform hand calculations and verify. Also added approaches 1 and 2 for top event risk calculation, 1 being convolving basic event fragilities withazard first and propagating risks through the fault tree. In approach 2, the fragilities are propagated first using the Min-Max method and the top event fragility is convolved with the hazard to calculate risk.