Pasar al contenido principal

Resultados relacionados

Resultados totales (Incluyendo duplicados): 2
Encontrada(s) 1 página(s)

Dipòsit Digital de Documents de la UAB
Publicación Artículo científico (JournalArticle)

Adapting WWTP performance to seasonality and climate change, Temperature-driven strategies for enhanced nitrogen removal

  • Revollar, Silvana|||0000-0001-9869-6437
  • Meneses, Montse|||0000-0002-5846-2676
  • Vilanova, Ramon|||0000-0002-8035-5199
  • Francisco, Mario|||0000-0002-6505-0936
  • Vega, Pastora|||0000-0002-3504-2273
A Temperature Driven Control Strategy that combines modifications on DO set-point, external carbon dosage and Solids Retention Time (SRT) is proposed to address the effect of seasonal and abnormal variations of temperature in the operation of Wastewater Treatment Plants. The medium-sized municipal WWTP, represented by Benchmark Simulation Model No. 2 (BSM2) that includes a Nitrogen Removal Activated Sludge Process (N-Removal ASP) with Dissolved Oxygen (DO) control has been selected as reference plant. The objective is to formulate a control strategy that improves the efficiency of the N-Removal ASP despite seasonal and unexpected variations of temperature of the influent, considering the trade-offs between temperature effects and control action effects on dynamic behavior. Simulations were carried out to evaluate the temperature and control actions effects on global average and dynamic behavior over a 30 days operation window, considering the BSM2 range from 10°C to 20 °C. A comprehensive analysis is carried out to look for potential solutions to improve performance at different temperature scenarios through the proper adjustment of DO set-points, carbon addition and sludge retention time (SRT). At low temperatures, the control actions are selected to improve N-Removal balance, since nitrification and denitrification rates are slowed producing violations of effluent quality limits in the presence of strong disturbances in influent flowrate. At high temperatures operational goals are to reduce the electricity consumption and the use of chemicals (external carbon), maintaining satisfactory performance. The Temperature Driven Control Strategy with different settings was tested considering a one year operating period. The proposed strategy address satisfactorily the seasonal and unexpected changes of temperature modelled by the annual BSM2 influent profile, attaining an appropriated trade-off between effluent quality and electricity consumption. The violations of required total Nitrogen limits are reduced up to 50 % in frequency and 33 % in length, while violations of required ammonium limits are reduced up to 36 % in frequency and 40 % in length. Annual average ammonium levels are reduced up to 19 % and annual average total Nitrogen is reduced up to 11 %. All these improvements are attained together with a slight decrease on electricity consumption. The proposed strategy is evaluated in a simulation platform, but it can be the starting point for the formulation of temperature driven advanced control strategies able to face climate change effects in the operation of for full-scale WWTPs.
Dipòsit Digital de Documents de la UAB
Publicación Artículo científico (JournalArticle)

Intelligent Control of Wastewater Treatment Plants Based on Model-Free Deep Reinforcement Learning

  • Aponte-Rengifo, Oscar
  • Francisco, Mario|||0000-0002-6505-0936
  • Vilanova, Ramon|||0000-0002-8035-5199
  • Vega, Pastora|||0000-0002-3504-2273
  • Revollar, Silvana|||0000-0001-9869-6437
In this work, deep reinforcement learning methodology takes advantage of transfer learning methodology to achieve a reasonable trade-off between environmental impact and operating costs in the activated sludge process of Wastewater treatment plants (WWTPs). WWTPs include complex nonlinear biological processes, high uncertainty, and climatic disturbances, among others. The dynamics of complex real processes are difficult to accurately approximate by mathematical models due to the complexity of the process itself. Consequently, model-based control can fail in practical application due to the mismatch between the mathematical model and the real process. Control based on the model-free reinforcement deep learning (RL) methodology emerges as an advantageous method to arrive at suboptimal solutions without the need for mathematical models of the real process. However, convergence of the RL method to a reasonable control for complex processes is data-intensive and time-consuming. For this reason, the RL method can use the transfer learning approach to cope with this inefficient and slow data-driven learning. In fact, the transfer learning method takes advantage of what has been learned so far so that the learning process to solve a new objective does not require so much data and time. The results demonstrate that cumulatively achieving conflicting objectives can efficiently be used to approach the control of complex real processes without relying on mathematical models.