Análise dos perfis de consumo domiciliar de água e efetividade das campanhas de gestão de demanda

Abstract

The residential use of water has played a critical role for many cities, mainly due to the reduction in quality and supply combined with the increase in demand in recent years. In this sense, the research carried out proposed to analyze residential water consumption according to end uses, identifying the impact and continuity of demand management campaigns to rationalize residential water consumption, as well as identifying possibilities for replacing treated water with alternative sources based on measured consumption. Therefore, the research was carried out in the city of Goiania, more specifically in the Jardim Nova Esperanca. After the interview process, filtering criteria, and withdrawals, 60 families were monitored during the research, and of these, 15 received telemetric meters to monitor the detailed consumption of water. Based on consumption data obtained from homes, especially those with smart meters installed, a campaign was developed to encourage optimized consumption using concepts of social marketing, economics, and behavioral science. Through drawing lots and consumption criteria, families were divided into two groups of equal size; the intervention group (IG), which suffered interference through the incentive campaign, and the control group (GC), which served as a parameter for the changes observed in the first group. Data collection for the development of the campaign took place during the first five months of follow-up, while verification of the results took place in the last seven months, totaling one year of data collection. The identification of the consumptions obtained in the telemetric meter used machine learning concepts, in particular artificial neural networks, and also the use of a gradient vector to separate events and episodes. Thus, it was observed that in the month following the campaign, there was a consumption variation of -5.67% in the GI, while in the CG there was a variation of 5.34%, and in the sequence of months there was a maintenance in the reduction initially observed in GI. This can be said in relation to the consumption of showers, toilets, and faucets, with no such behavior being observed in the washing machine, swimming pool, and external uses. Regarding alternative uses, there was a reduction of up to 61.57% in the consumption of treated water in substitution for greywater and rainwater, with the possibility of a significant reduction in the concessionaire's billing if applied on a large scale. Thus, the potential of using consumption information in the development of campaigns for demand management is demonstrated, allowing the development of important tools for the management of water resources.