Styrning av biologisk kväverening anpassat efter tidsvarierande elpris

The electricity demand of a waste water treatment plants follows the diurnal pattern of society, and this generally leads to higher demand when the market price is high. The possibility to adapt the operation after price variation has been known since long, but few studies have been published. It ha...

Full description

Bibliographic Details
Main Author: Sund, Johan
Format: Others
Language:Swedish
Published: 2019
Subjects:
Online Access:http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-388459
Description
Summary:The electricity demand of a waste water treatment plants follows the diurnal pattern of society, and this generally leads to higher demand when the market price is high. The possibility to adapt the operation after price variation has been known since long, but few studies have been published. It has been suggested that the influent can be redistributed using an equalization basin, and one study showed 16 % reduction in cost with equalization to constant flow. Oxygen supply by aeration uses the major part of electricity, and adaptation of aeration intensity has also been suggested. However, this requires respect for effluent limits, especially for nitrogen, as larger plants are often equipped with nitrogen removal. In this study, optimal control of aeration was used to evaluate the potential of adapted aeration. Use of an equalization basin was also studied. A reduced version of Benchmark Simulation Model no. 1 was used, with only one basin. Aeration was optimized for minimal cost given a price profile for 24 hours, under a constraint on ammonia discharge. Cost was reduced with 1-2.5 % compared to energy-optimal control. Constant flow equalization showed an energy reduction of 2.5-12 %, and a cost reduction of additionally up to 5 %. Control adapted after price gave another 1-3 % savings. The nitrification process is sensitive to oxygen and ammonia concentration. This makes it difficult to redistribute nitrification over the day, especially with a one basin model. It is therefore motivated to study a model with more basins.