Risk assessment of water quality using Monte Carlo simulation and artificial neural network method.
Recreational outbreaks associated with sprayparks are well recognized, and may be partly due to the engineering designs used for their water reclamation systems are problematic to control. This work is based on an outbreak of cryptosporidiosis linked to a spraypark in New York State, where it was determined, specifically that the spraypad (the main attraction) was the primary exposure point. We first determined the likely dose the spraypad users were exposed to, then modeled the efficacy of the treatment system and used this to inform a Monte Carlo method to estimate the probability of infection and illness for the users of the spraypad. The current treatment system which consists of; two holding tanks, a dual media filter and chlorine injection as well as two design change recommendations were modeled using three independent Markov chain models. Within the current treatment system design the receiving tank for the treatment train is also connected with a second pipe to the spraypad used to deliver the return (treated) water, this return pipe is acting potentially as a bypass for the treatment train. Based on the risk assessments performed it is recommended that the bypass pipe be removed from the treatment system since in doing so the probability of infection and illness were reduced appreciably. Secondarily including an ozone contactor was shown to slightly reduce the risk further and provide a multiple barrier.