Texas City Showed We Still Don't Learn from Experience

By Andrew Bond, Industrial Automation Insider

Jul 06, 2010

Delivering the keynote address at Curvaceous Software's customer forum in the U.K. last month, former Abnormal Situation Management (ASM) Consortium program director Ian Nimmo showed how the same conclusions had been drawn from the investigations into the causes of the Texaco Milford Haven explosion in 1994 and the BP Texas City accident 11 years later. Nothing had changed – the same culture and the same control room and alarm management failings had created the same problems for operators and the alarm system and operator HMI had both been identified as significant contributing factors to both accidents.

Nimmo is now president of User Centered Design Services (UCDS), the company he founded to focus on implementing best-practice solutions for reducing the frequency and severity of abnormal situations. He had been introduced to Curvaceous managing director Robin Brooks at the Control Room Conference in London last December where Brooks had presented a paper on Curvaceous’ newly introduced Visual Explorer 2.5 product. That paper had prompted Nimmo to comment that "This was the first alarm management paper in the last ten years that got me excited, as Dr Brooks explained how alarms can achieve management objectives, how to understand alarm settings in a new light, and how to change them correctly."

Good Practice
During his time with the ASM Consortium, Nimmo had encouraged and supported the production of the EEMUA191 Alarm Systems Guide to Design, Management and Procurement in 1998, while NAMUR published an alarm management guideline in 2003 (NA102) and the ISA Standard 18.02 is currently nearing completion. All these guides present good engineering practice, but none is compulsory, although most process plant operators have an internal corporate mandate to improve their alarm management systems. In the U.S., OSHA outlines certain critical minimal requirements for alarm systems in CFR29, regulation 1910.119, Process Safety Management of Highly Hazardous Chemicals. Moreover, public pressure and labor unions are demanding stronger action. Following the Texas City accident, OSHA announced a plan in 2007 for extensive monitoring of refinery compliance, using 300 inspectors in a role similar to the U.K. HSE yearly inspections of refineries.

While acknowledging all this activity, Nimmo stressed in his address to the Curvaceous forum that alarm management is just one aspect of control room management. Based on ASM studies and in conjunction with other authors, he has now produced a further guide, entitled the High-Performance HMI Handbook to address the perceived failings of current HMI displays. Based on experience, the UCDS approach is to look at the whole control room environment from the operator's viewpoint. The company provides the specialist skills needed to improve control room management and systems by re-examining operator training, traffic into and through the control room and other procedures.

Recent projects by UCDS have addressed many of the HMI issues covered in the HMI Handbook, using screens designed to improve operator situation awareness. Many of the developments incorporated use the trend and analog displays that were originally standard in control rooms, with color used on HMIs only for specific events or alarms.

Exciting Possibilities
Nimmo places great emphasis on the information that should be available from trends in data and operational history in establishing effective plant control and believes that exciting possibilities could be created by incorporating into future USDC control system projects Curvaceous Software's geometric process control techniques, which are themselves based on process performance history.

An offshoot of the consultancy activities of Process Plant Computing Ltd, the company which developed the original Geometric Process Control (GPC) concepts on which it was based, Curvaceous Software has grown considerably under Robin Brooks since its establishment as a commercial software provider in 1998.

However, with the development of the Web, the name is in danger of becoming more of a liability than an asset. Web searches for "Curvaceous control" could lead potential clients into areas not necessarily deemed appropriate for a business or indeed a family environment, although our own one page Google search turned up nothing more salacious than "Voluptuos (sic) Nigella Lawson struggles to control her curvaceous figure," the aforementioned Nigella being both a celebrity TV cook in the U.K. and the daughter of former Chancellor of the Exchequer and now scourge of climate change fanatic, Nigel, now Lord, Lawson.

In a move designed to protect potential customers' blood pressure, a different company name and product identity was initially proposed for use solely in those areas, principally the U.S. and China, where the original name could prove particularly problematic. However, the proposed new name, C-Visual, presented even more problems, raising the possibility of a trademark conflict with Microsoft Visual and C++. As a result, the decision has now been taken to adopt a new overall company identity allowing for a global branding and internet presence. Consequently, from the end of April 2010, Curvaceous Software becomes, indeed in effect reverts to, PPCL with a new identity and logo and a new website at www.ppcl.com while Curvaceous Visual Explorer 2.5 becomes CVE 2.5.

Ian Nimmo's views on support for the operator are to a significant degree echoed in a new white paper from Longwatch president and former Intellution founder and CEO Steve Rubin entitled "Tuning the Forgotten Loop." Rubin's thesis is that, while operators are under ever increasing pressure from, for example, new Six Sigma quality demands, new technologies such as Smart Grid, increasingly complex HMI screens and insurance and regulatory issues, few tools are available to train or 'tune' them. His solution is to provide operators with visual feedback of their responses to alarms, incidents and accidents by combining video recordings of the plant and the operator consoles with data from plant historians to provide "instant replay" of an operator's actions or of the actions of an experienced operator handling the same situation. No prizes for guessing whose technology is to be used to obtain the video recordings. Download the white paper from http://longwatch.com/thedocs/Forgotten%20Loop.pdf

Show Comments
Hide Comments

Join the discussion

We welcome your thoughtful comments.
All comments will display your user name.

Want to participate in the discussion?

Register for free

Log in for complete access.


No one has commented on this page yet.

RSS feed for comments on this page | RSS feed for all comments