Energy efficiency optimisation of wastewater treatment: study of ATAD

The aim of this investigation was to minimise the energy requirement of autothermal thermophilic aerobic digestion (ATAD) systems while complying with treatment objectives. Due to the discontinuous, semi-batch nature of the ATAD reaction, these systems have to be optimised via dynamic optimisation. To this end, two dynamic ATAD models were developed: ATM1 and ATM2. These models have the novel and unique feature that they are capable of quantifying the two treatment objectives, i.e., stabilisation and pasteurisation. Thereby, they are also able to determine the minimum energy requirement needed to satisfy legal standards. Simulation studies were carried out to examine model behaviour of different ATAD systems at start-up and steady state operation. Overall, the behaviour ob- served in the simulations was in good qualitative agreement with the behaviour of full-scale plants. In an asymptotic analysis of the ATM1 model, the general, quali- tative structure of the solution was studied. The structure is composed of a number of regions which are limited by different factors: first by the availability of dissolved oxygen, second by readily biodegradable substrate, and nally by slowly biodegradable substrate. The structure of the solution can explain the qualitative behaviour of respirometric curves of ATAD systems. A global sensitivity analysis was carried out to identify the model parameters with the strongest in uence on energy requirement and plant capacity. In this context, it was found that reactor volume, aeration owrate, reaction time, loading time, and volume replaced after each batch were the most significant parameters. It was also found that the ATAD reaction is generally limited by the stabilisation process. A general relation of inverse proportionality was found between energy requirement and plant capacity, which holds for other wastewater and sludge treatment processes. The potential implications of this finding are yet to be explored. A model assessment of ATM1 was carried out, indicating a good qualitative agreement between data and simulations. Nonetheless, there exists room for improvement in quantitative terms. The optimisation problem was then formulated within the framework of the direct sequential approach and it was solved for a single-stage (CS1) and a two-stage (CS2) system. After optimisation, their energy requirement had been reduced by 23 and 18%, respectively. An even better solution (with 42% reduction) was found for CS2 by assuming a pre-dewatering stage and treating the remaining sludge in the rst-stage reactor. Finally, a tentative framework for dynamic optimisation of ATAD within the direct simultaneous approach was proposed. The optimisation problem was reformulated. Even though the problem was not solved, simulations studies pointed out some of the di culties of implementing and solving this problem and their potential solutions. In the future, the latter framework shall prove useful for the structural optimisation of ATAD systems. Optimisation is rarely used in wastewater engineering. Given the high, rising cost of wastewater treatment, optimisation should become the norm for design and operation of wastewater treatment plants.