PPCL Logo Process Plant Computing Limited

Tel: +44 1753 893090 |

LinkedIn Twitter

Insight from past performance

Discover why you have a process historian! You knew there was lots of new process knowledge representing improvement opportunity hidden in all that data and thought that you could extract it. Then you realised that engineers lacked the tools they needed so could only pick at a few highlights. CVE is the tool that you have been missing, and with CVE you will discover exactly why it was such a good decision to buy a process historian.  

Process plant data is different from the data usually explored with the traditional 'data mining' and 'predictive data analytics' methods that you may have learnt in college because it is highly correlated amongst hundreds or thousands of variables so there are many, many correlations between variables. That means the problem is not one of finding correlations but of recognising those that were previously unknown and have value amongst the many that you already know, or that are direct consequences of the underlying chemistry, mass, heat and momentum balances in your particular process.

Correlations in themselves don't identify cause and effect. That requires the engineer's process knowledge, so the method needed also has to be quick and straightforward to use and explain to others for busy engineers with many other tasks to accomplish during their working day.

Process plant data is also often highly non-linear. Non-linearities can easily be seen in CVE, often for the first time, which adds to its uniqueness and demands the ability provided by CVE to easily separate out individual modes of non-linear behaviour.

C Visual Explorer (CVE)
 provides a graphical representation of the values of many hundreds to several thousand process and result variables using a parallel coordinates graph tailored by us as chemical engineers to the particular needs of process plants. The parallel coordinates graph is a coordinate transformation from n-dimensions (each variable is a dimension) to 2 dimensions and many other mathematical properties. This unique visual method allows engineers to quickly see and extract process insight from the data already collected by your historian without involving any mathematics on the part of the user. CVE can be used for continuous and batch processes and even for mixed batch/continuous processes. CVE is fully supported and is continuously being developed.

Using CVE you can:

tik Understand your process and how it can best operate by combining process historian data with data from other repositories such as laboratory analyses, emission measurements, raw material assays and more.
tik Optimise your operations without changing how you work by understanding what an Operating Window really is and what we mean when we talk of a 'consistent' window.
tik Optimise even more, but disruptively, by finding the Operating Envelope required to achieve a particular set of business objectives and using the Operating Envelope in real time (requires CPM also) to implement 'boundary control' to supplement existing process controls.
tik Find better operator alarm limits in a fraction of the man-days you take with today's methods and have immediate alarm performance predictions available as you interactively change alarm limit values before and during the alarm review meeting. Our methods will take you all the way to predictive alarming (requires CPM) if your ambitions extend that far.
tik Find the Operating Windows and Envelopes that show you where to operate to meet all your KPIs, not just the ones you can measure in real-time.
tik Find which variables are contributing most to variability.
tik Use CVE on a daily basis for process stewardship by your process engineer. Have him or her share the experience with the unit operations team. They are best placed to understand and explain what is different about today's operations and why, and in much the best position to make immediate corrections. You will find average efficiency increasing and variability decreasing.
 
Your historical operating data and offline quality measurements contain incredible amounts of information about how your process behaves that you really haven't been able to extract until now. By unlocking this information you can regain control, reduce waste, increase yields, improve product quality, control your emissions and ultimately reduce costs!

But How?

CVE uses a unique form of geometry to convert historical data into a single visual summary. The technology also enables you to select periods when top quality outputs were achieved and then control your process to only mimic these conditions. The result? A better process you are in control of!

What's New in CVE 2.7?

Improvements to the Cluster Variable

CVE implements multi-variable clustering with no limit imposed upon the number of variables to be clustered. The handling of over-lapped clusters has been altered so that points are assigned to the largest existing cluster meeting proximity requirements rather than the first cluster found. The time to initialise the cluster variable has been dramatically reduced. Clusters are numbered starting with zero and the created query description text has been altered to be consistent. New examples have been added.

Integration Variable

New examples have been added showing how to use the integration variable to produce totalisations for different time periods which is especially useful in monitoring and reporting Environmental KPI Limits expressed as total emissions per month or per year. Guidance text in the Edit Box has been improved and the Integration Variable now appears in the User Guide Table of Contents.

Capture Image of Pareto Plot

Capture Image now captures all of a very wide Pareto Plot as a series of panels similarly to how it already captures images of very wide parallel plots.

More new features:

  • Alarm Forensics extends Alarm Analysis and Alarm Performance Prediction to individual variables for detailed analysis of alarm problems, alarm floods and finding consequential alarms.
  • Box Variable for prediction of Yields and identifying the most important variables for reducing variability.
  • Interface to Honeywell PHD 310 for retrieval of historical data directly without intermediate Excel files.
  • CVE 2.7 has been tested with 1,000 alarms and variables. Most features work well with even more.
  • Pareto plots with time control for comparing sets of limits.
  • Time-trend plots with automatic compression to maximise visual resolution.
  • Everything synchronizes automatically: Parallel plot, Pareto plots, Scatter plots, Distribution plots, Time-Trend plots, Queries and Calculations.
  • Window Manager for controlling available windows improves work with many variables.
  • Find and Locate functions operate across Parallel and Pareto plots for superior navigation.
  • Improved window positioning for trouble-free switching and image-capture between multi-monitor desktop configurations.
  • Windows 10 is now supported except for its touch-screen features, which are not used by CVE. Windows 8 (excluding touch features) and 7 remain supported as well as XP, Vista, and Microsoft Windows Servers from 2008. Support for Windows XP and Windows Vista will be discontinued in CVE 2.8. “Supported” means that new software fixes will be created as necessary for CVE problems that can be reproduced in an environments with Windows running natively on the PC hardware.
  • Reduced memory usage and increased performance.
  • Edges colour-coded and directional for effective use of many simultaneous queries.
  • New model data interchange with C Process Modeller (CPM)
  • New licence management software for CVE Network Edition adds functionality.
  • User Guide and online help fully updated.

What's Next?

Watch the video on the right and, for a more in-depth look, book onto one of our introductory webinars: www.ppcl.com/webinars.  For further details, please email This email address is being protected from spambots. You need JavaScript enabled to view it. or call +44(0)1753 893090.
 

Register Now!
PPCL Webinar: Alarm Rationalization
7th & 8th June 2017

PPCL © Copyright 2011-2017. All Rights Reserved. | Terms of use | Privacy Statement | P.O. Box 43, Gerrards Cross, Buckinghamshire, SL9 8UX. UK.