Capturing Connectivity and Causality in Complex Industrial Processes

Yang, Fan; Duan, Ping; Shah, Sirish L.; Chen, Tongwen

doi:10.1007/978-3-319-05380-6

Cited by 56 publications

(40 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…However, there is no causality between F 2 and F 3 , since by ruling out their common cause, F 1 , F 2 and F 3 are both independent. (c) Causality; adopted from Yang et al [7] From the illustration above it can be concluded that correlation is a necessary (not sufficient) condition for causality. Beside correlation, some additional conditions are required to imply a causal relationship, such as connectivity, responsiveness, and the direction of the relationship between two process variables [7].…”

Section: Inter-relationships Of Process Variablesmentioning

confidence: 98%

“…Connectivity, causality, and correlation are three important attributes which are used to describe such inter-relationships [7]. The illustrative example in Figure 1, adopted from Yang et al [7], is provided to facilitate better explanation of the physical interpretation and practical use of these different concepts. As shown in Figure 1, a liquid hydrocarbon feed stream is fed into a distillation column.…”

Section: Inter-relationships Of Process Variablesmentioning

confidence: 99%

See 1 more Smart Citation

Multivariate probabilistic safety analysis of process facilities using the Copula Bayesian Network model

Hashemi

Khan

Ahmed

2016

Computers & Chemical Engineering

View full text Add to dashboard Cite

Section: Inter-relationships Of Process Variablesmentioning

confidence: 98%

Section: Inter-relationships Of Process Variablesmentioning

confidence: 99%

Multivariate probabilistic safety analysis of process facilities using the Copula Bayesian Network model

Hashemi

Khan

Ahmed

2016

Computers & Chemical Engineering

View full text Add to dashboard Cite

“…Within the process engineering and control communities there are a number of approaches to cause and effect analysis and disturbance diagnosis (for a review see Horch, 2007, andYang, Duan, Shah andChen, 2014). These approaches include the use of graph theory to model causal and connectivity relationships using process and instrumentation diagrams (P&IDs) and other drawing information (e.g.…”

Section: Introductionmentioning

confidence: 99%

Process Disturbance Cause & Effect Analysis Using Bayesian Networks

Leng

Thornhill

2015

IFAC-PapersOnLine

View full text Add to dashboard Cite

Process disturbances can propagate over entire plants and it can be difficult to locate their root causes from observed effects. Bayesian Networks offer a way to represent unit operations, processes and whole plants as probabilistic models which can be used to infer and rank likely causes from observed effects. This paper presents a methodology to use deterministic steady-state process models to derive Bayesian Networks based on alarm event detection. An example heat recovery network is used to illustrate the model building and inferential procedures.

show abstract

“…Connectivity and causality have been concepts proposed for describing dependencies between units and variables [1]. Several authors have been applying new methods that make it possible to extract process topology describing causality relationships from process knowledge and process data.…”

Section: Introductionmentioning

confidence: 99%

Integration, navigation and exploration of plant topology networks using the property-graph model

Romero

Thornhill

2014

2014 IEEE/SICE International Symposium on System Integration

View full text Add to dashboard Cite

Abstract-Understanding the process topology is essential for many process systems engineering activities. Previous works in this area have explored the extraction of connectivity and causality from different sources by using structural and process data. A few works, however, have focused on how to integrate, store, and visualise connectivity models, which are becoming available in the literature. This paper proposes a method for integration, navigation and exploration of topology models in an efficient way, which can be applied to the study of connectivity and causality in chemical processes. I. INTRODUCTIONDisturbances in continuous processes commonly travel along the product stream through different propagation paths affecting the process performance. Abnormalities that propagate plantwide include oscillations caused by sticking valves, disturbances produced by interacting controllers, instabilities in the systems such as slugging flows, and vibrations caused by interacting units (for example in a turbine-generator train).Since large-scale chemical facilities are highly integrated systems containing recycle flows and utility systems, units are not independent anymore. Connectivity and causality have been concepts proposed for describing dependencies between units and variables [1]. Several authors have been applying new methods that make it possible to extract process topology describing causality relationships from process knowledge and process data. The knowledge from process topology has many applications, for instance in the field of plantwide analysis, where the use of process connectivity can enhance the isolation and diagnosis of the root-cause of plantwide disturbances [2], [3]. There is also evidence of use of topology in other areas such as alarm rationalization [4], risk assessment [5] and control structure design [6]. It seems now that mining the data and knowledge about topology (connectivity and causality) is a largely solved problem. What becomes challenging now is to integrate all these insights into a single topology model and make it possible for engineers to use it for exploration, navigation and more effective visualisation.This work is supported Marie Curie FP7 project ENERGY-SMARTOPS, Contract No: PITN-GA-2010-264940.David Dorantes Romero is with ABB Technology and Innovation, Norway, Ole Deviks vei 10, david.romero@no.abb.com. He is also part-time PhD student at the Department of Chemical Engineering at Imperial College.Nina F. Thornhill is with the Centre for Process Systems engineering, Department of Chemical Engineering, Imperial College London, London SW7 2AZ, U.K.The main contribution of this paper is the proposal of a method to integrate process connectivity and causality information into a plant topology network, making use of interaction and techniques to facilitate its navigation and exploration. Examples will also be given based on a novel prototype called Topoviz™, currently under development. II. STATE OF THE ARTInformation about the topology can be captured from modelbased m...

show abstract

Capturing Connectivity and Causality in Complex Industrial Processes

Cited by 56 publications

References 0 publications

Multivariate probabilistic safety analysis of process facilities using the Copula Bayesian Network model

Multivariate probabilistic safety analysis of process facilities using the Copula Bayesian Network model

Process Disturbance Cause & Effect Analysis Using Bayesian Networks

Integration, navigation and exploration of plant topology networks using the property-graph model

Contact Info

Product

Resources

About