Thursday, August 20, 2015

The need for new employees are nothing new - but it does influence process safety

It was a pleaant surprise to see Hydrocarbon Processing address the management of organisation change (MOoC) in the article "Update: Personnel changes introduce new considerations for PSM" by R.L. Gather in the June issue. I first encountered MOoC at a Technical Steering Committee (TSC) meeting in EPSC auspices, when we had to decide which project to priorities for management approval of funding more than 10 years ago. One of the proposal was for the development of standard procedures for personnel changes. However, although there were several in favor, the TSC decided, that this was not a technical issue, and hence the proposal was dropped from the list.

Those in favor at the time argued that MOoC was indeed relevant, because nowhere else in the organisations were the process safety impact of the changes listed in table 1 of Gather's article being considered. What do you think? Should the impact of personnel changes be analyzed for their possible process safety impact prior to implementation?''

I was a control engineer at a major Canadian site, when two of our control engineers at another unit quit to work for another company. Their quitting resulted in me being moved almost immediately to the unit they used to work at. I was a cool move for me - in retrospect - because the unit I moved to had the operators with the most positive attitude to testing new control strategies and even suggesting improvements to existing ones. At the time my move was a result of the unit being left with just one control engineer, which at the time was considered to thin coverage. This was back in the eighties, when call from headhunters was almost a weekly occurrence for control engineers at this site.

A related issue is addressed in the article "Overcoming the challenges of the 'great crew change' " by tech editor M. Rhodes. That is the eminent retirement of the post WWII baby boomer generation.  When I think back this was also a hot issue when I joined the industry about 35 years ago. So what has changed? I am really not sure.

Well generations change. Many of the people, whom I started by carrier with in the HPI still work or have recently retired from the company we all worked at 35 years ago. 35 years ago people joined a company an worked there almost for life. From talking to my now grown up kids, whom all four belong to the Millenials none of them expect to work for the same company all through their carrier. Even some of the people I studied engineering with in the early seventies today work for the same company, which they joined after graduation. There are naturally exception, but very few have had more than five jobs. So today companies cannot expect people to be with them for their whole working life. That has changed!

I have also noticed, that for the Millenials work and life mix much more. They are not eight to five workers.

However, I think one thing have not changed. When you get you first job, many would argue, that university have not prepared you properly for the job. I think that also was the case 35 years ago. But suddenly we realized that university problems solving skills could easily be converted to work places problems solving skills. Of course there were people to get to know and work with. But so there was in your university courses. You arrive at your first job loaded with the latest problem solving skills from university, and within a relatively short time these skills are applied in the new corporate environment.

I think one of the misconceptions is that new graduates are filling positions of retiring employees on a one to one basis. That was definitely not the case when I entered the HPI, and I don't believe it is today either.

Nonetheless the HPI as pointed out in the article has a challenge to attract a sufficient number of new graduates, but that in my opinion is due to the image of the HPI. By many graduates it is seen as a dirty and polluting industry. That was also my view until I have a chance to visit Imperial Oils Edmonton Refinery, while I studied at University of Alberta. You could walk around the refinery in a white shirt, and it wouldn't get dirty! (Of course today, that behavior would be considered unsafe and not allowed.)

Mittvakkat icebreaker photographed August 15th, 1933 and again Juli 31st, 2010
So in order to attract a sufficient number of new employees I think the HPI must changes its behavior. That means taking process safety much more seriously so we don't see headlines in the news such as after the recent explosion at ExxonMobil's California refinery or the explosion at a warehouse a few days ago. That also means taking sustainability more seriously. Fortunely resent announcements by Dow Chemical points in the direction. That also means talking climate change seriously. I think the two pictures above from Greenland says it all.

Wednesday, August 19, 2015

Cloud Computing will NOT revolutionize process safety!

I was a bit surprised reading the title of the article by J.Lucas and S.Whiteside in the June 2015 issue of Hydrocarbon Processing. It was "Cloud Computing: The next revolution in process safety!". The HP editor was a bit more modest on the cover of the issue by stating "PROCESS SAFETY Using 'the cloud' can improve programs and reporting".

Illustration courtesy of Wikipedia.
Let me make it clear: In my view the term "cloud computing" is a marketing gimmick created by the IT-industry. It covers the fact that in stead of running on servers in your own data-center your applications run on a service-providers hardware somewhere in some country. The service provider can be companies such as Google, Amazon, IBM and other who allow you to access their hardware and dynamically change the amount you use by the day, hour or even minute. The new thing is that you can configure and access the hardware you rent a right to use through the internet using a secure channel. Some providers even call the process of configuring such systems for orchestration. Once the applications have been deployed on the providers hardware you can grant others access to them using the same channel used to configure them.

However, that would also be possible using your own hardware in your own data-center."The cloud" in my view is just a term which indicate you have less knowledge about where your application and data physically are located (I know the IT-industry also use concepts such as private cloud, hybrid cloud and public cloud, but that just has to do with whom you are sharing the hardware with).

To the best of my judgement Lucas and Whiteside are not subject matter experts in process safety. That is clear already from Table 1 in the article with the heading "History of Process Safety Management", which starts with the 1984 MIC release in Bhopal and apparently end with the 1998 publication of  the standard IEC 61508 according to the authors. Process Safety Management in my view started when ICI began using a tool called hazard and operability (HAZOP) analysis to identify safety issues in their new plant designs before the facilities were constructed, and shared this idea with the world. And properly the most significant recent event for process safety management was the publication in early 2007 of "The Report of the BP U.S. Refineries Independent Safety Review Panel" (.a.k.a. The Baker Panel Report). This report has indeed created a revolution in how company boards address process safety management.

Illustration courtesy of Wikipedia.
The article by Lucas and Whiteside describe how process safety engineers started using desktop applications, such as spreadsheets, in the eighties and then moved to document management systems in the nineties (I actually experienced the arrival of the IBM PC as a tool for control engineers in the mid eighties, but even at that time it was a tool to run GUI applications on. Such applications could not be run on the mainframes of that time. However, the mainframe still served as a file repository for the PC).

Lucas and Whiteside further state, that the lack of standardized file naming conventions made it difficult to identify relevant files in e.g. a document management system. I certainly agree with that. However, "the cloud" does not introduces any standardized file naming convention. However it is most certainly true, that cloud based technology such as Google Apps for Work, Microsoft OneDrive and Drop Box have made it easier to share files across organizations and even highly distributed groups (I am currently involved in a commercial project, which use Drop Box for file sharing since we are currently a rather small group and it was easy to set up). However such sharing was possible possible through the mainframes using character based terminals in the nineties.

Further on Lucas and Whiteside describe what they call safety management with live data. However, without access to the plant data historian in the DCS there will be no live data. As far as I know such data sharing directly from the plant data historian in the DCS is still rather experimental technology. Without such live data access it is difficult to see how the cloud can provide a plant safety boost or a revolution in process safety. Such a revolution would in my view need to access real time plant data.

In my view a key to improving process safety is to develop tools to assist the plant operators in responding to deviations or even potential deviations from normal operations. That is an active area of current research in which I am a bit involve. You could argue, that hence I am biased.

What do you think?