IMPROVING DECISION MAKING IN WATER PLANT OPERABILITY THROUGH BAYESIAN BELIEF NETWORKS
How to get more from existing instrumentation data to achieve better process performance
T Trinh, C Pelekani, G Leslie, P Le-Clech
Publication Date (Web): 29 June 2017
The increase in stringency of water quality requirements in Australia has driven the need for improved data collection and process monitoring practices at water treatment plants. Despite the clear benefit of online monitoring for risk reduction and improved compliance, obtaining the full effective value of large volumes of data created by online instruments is still an ongoing challenge. This study aims to apply a qualitative risk assessment tool called Bayesian Belief Network (BBN) to expand the use of historical data for improving decision making in water treatment plant operability.
BBN is a graphical model that represents a set of variables and their probabilistic dependencies. In BBNs, variables are represented by nodes, and the relationships between variables are represented by directed arcs. Quantitatively, these relationships are expressed in conditional probability tables. BBNs offer numerous benefits for modelling complex systems but their applications in water treatment systems are still very limited. A few studies have investigated the application of BBNs in diagnosing upsets in lab-scale wastewater treatment systems. However, application of BBNs based diagnosis systems for full-scale water treatment processes has not been reported. In this study, based on available online turbidity data and related operational inputs from a filtration process of a full-scale water treatment plant, BBNs were developed and validated, which could determine the probability of several possible causes and the corresponding corrective actions for given high filter outlet turbidity readings. Such quantitative statistical information from the models could help operators to decide on the appropriate courses of action when facing different situations at the plant, and thus improve day to day decision making during ‘out of normal’ operations.
Applying BBNs in this environment presents a number of challenges, in particular, obtaining the appropriate dataset and information for the model development and validation. On one hand, plants with sound management strategies and good monitoring records usually have little “out of normal operation” incidents. Although full dataset could be obtained from these plants, the variation of the data results in few relevant incidents, and limit the full development and validation of comprehensive BBNs. On the other hand, plants featuring a wide range of “out of normal operation” incidents, are usually not well monitored and lack of appropriate dataset. The full benefit of BBNs therefore appears to be limited to plants featuring sufficient “out of normal operation” incidents and appropriate amount of monitoring data and operational and maintenance records. As a result of this study, it has been demonstrated that better decision tools can be developed from historical data. In addition, BBNs can be a good training tool for new operators as the models can offer better understanding of the links between different parameters in the processes. Furthermore, BBNs can also serve as a complementary strategy to existing management strategies of the plant to improve the reliability of its processes.
Click here to read the full paper
Please note: You need to login to access this member-only content.
Not a member? You can purchase the paper in our Online Document Library.