by A digital twin for a treatment plant – prediction meets reality!
When we live in a world with advanced technologies as simple as Alexa helping you play a music track, to a Tesla driving your vehicle making daily human lives much easier, have you ever wondered how technological advancements are helping in the environmental sector? in case of wastewater treatment plants? Of all the recent development in the field, digitalisation and use of artificial intelligence are of greatest impact.
Model based studies on a treatment plant will not only save energy, money and time but also give information about the unmeasurable unknowns with the help of parameters that can be measured. Digital twin is a model platform of growing popularity. Though the concept is not novel and has been in existence for more than a decade now, the term ‘Digital Twin’ essentially means a digital replica of the treatment plant, representing all aspects of the plant including physical infrastructure and processes in real time. It provides the plant performance on a real time basis and a predictive future performance through a live two-way digital link between the digital and physical worlds.
First step is to develop the framework of digital twin with the physical (dimension and location of different compartments), operational (loading rate, temperature and flow rates) and process (physical, chemical and biological removal) aspects. Data input is the second stage, digital twin works on live data updated to the model from the real plant by manual collection or integrating devices like sensors, meters, etc. Data analytics should be done with utmost care for an accurate digital twin. In a predictive analystic approach, future performance of the plant is predicted from past behavioural data collected under different situations. Digital twin thus works on a data driven decision making process by utilising historical and real time data.
Isn’t it amazing to know how a treatment plant will work a few weeks later or in an extended future? A lot of planning and optimisations can be done for a better future functioning of the plant based on the predictions made. Many unforeseen events affecting the plant performance can be discovered earlier and what more! I leave it open here for you to think about!!
