Today’s young process engineers have tools available to them that we “old timers” only dreamed about. Although smaller and some medium sized companies will not have the items listed below, most large refining and or petrochemical operations have the following:
- Real-time and historical plant operating data available on his/her desk top computers;
- Real-time and historical plant laboratory data available on his/her desk top computer;
- Access to the plant DCS control schematics showing current operating details (though no ability to change set points, etc.);
- Access to powerful process simulation software;
- Up-to-date (mostly) P&ID’s and unit piping/structural drawings available with the click of a mouse.
The list of available resources to today’s younger process engineer without leaving his or her office goes on and on. Couple this with the trend to large centralized control rooms for safety and efficiency reasons and you easily end up with one large negative to all this progress. There is very little apparent reason and also a bit of a hassle to interact with the unit outside operators.
Compare this with my generation (60’s and 70’s) doing plant process engineering work.
- My initial calculation device of choice was the slide rule. I thought I’d died and gone to heaven when Hewlett Packard introduced their first calculator.
- Much of my “data-acquisition” was obtained using the pencil/clipboard method getting local readings and “averaging” strip-chart record data in the control room.
- Laboratory data was not nearly as extensive as it is today and many results were a function of the lab technician running the test.
- Process simulations were normally equipment specific (e.g. distillation) and were generally Fortran programs running on a main-frame computer system not located on-site.
- We relied on API correlations to infer composition and physical properties of many process streams.
The move to large consolidated control rooms and distributed digital control systems (DCSs) was just beginning. I was personally involved with installing the first Foxboro DCS on a reforming and petrochemical unit.
This transition from clipboard to chip board gave us old timers some advantages.
- We were happy to get even rudimentary data at our desk but we knew which readings could be flakey, false or not reliable.
- We were able to see the control system graphics but we knew which lines were missing and which operations were not being done as show on the graphic.
- Most of all, we were interacting with all relevant unit personnel; a priceless source of good operating information.
Where is this going? My job as an aged “sage” consultant is to drag the newly minted process engineers out from behind their desks and out on the unit such that they get a feel for the plant and to get them to interact with the unit personnel.
Case in point – at a large seaport refinery there was a problem with excessive amounts of hydrocarbon liquids going to the uncontaminated sewer system. I was asked by the process engineering manager to assist a young process engineer to identify the source of the leak and come up with a fix.
Since I had just come on board, I was viewed with some suspicion (perhaps not unwarranted?) In our first meeting I asked him what he proposed to do? He stated he wanted to get the refinery laboratory to identify the components in the sewer sample (correct) but they would be requested to use techniques and equipment unavailable to the refinery lab and the sample would have to be sent to the company’s more sophisticated laboratory in another location. Estimated turnaround time for this to happen was from 3 days to a week with a lot of “politicking” necessary to achieve the priority needed.
I suggested that we do this the old fashioned way when there were real process engineers following units. [I did not get nearly the pushback I had expected!]
We got the lab to determine the sewer sample API gravity and run a [simulated] D-86 distillation.
Next we used techniques and correlations found in the API Data Books section on characterization to calculate a mean average boiling point for the sample.
Using additional graphical correlations listed in the API Data Book we determined:
- The molecular weight;
- A “Watson” or UOP K factor.
This established that the material leaking into the sewer was diesel. Upon this revelation the youngster said that he would write a letter to the Operating Department to get the unit operators to check their equipment for leaks!
Nice try, but it will never happen in your lifetime. Get your PPE together, 4 or 5 sample bottles from the lab, some good string and a knife; we are going on a sewer leak hunt! He asked how this was going to work? I said we sample all [uncontaminated] sewers discharging into the waste water treating facility until we find the culprit. Then we sample all upstream sewers discharging into that sewer and repeat that process until we find the culprit. At some point we will find the sewer box that is laden with diesel and all upstream sewer box feeding this sewer box are clean. Then we have to find the source of the leak into that sewer box.
We used this procedure along with me showing him how to “make” a sewer sampling bottle rig and we found the culprit sewer box in about one and a half hours. It was 100 feet away from a diesel hydro-treating unit’s contaminated sewer box. Hmmmmmm.
The youngster then asserted that it was impossible for the hydro-treater to be associated with our problem because the hydro-treater contaminated sewer was (should not be!) not connected to the uncontaminated sewer and besides no refinery print showed the connection. Hmmmmm.
So I suggested that we talk to an operator that looked a lot more like me than him and see if he could shed any light on the subject.
Sure enough, during a unit T/A some years back, a bootleg line (with a valve) was installed to connect the two sewers to get around a problem associated with the T/A. After the T/A was completed the valve was shut and line buried. No one was the wiser. But it is not on the print protested the youngster. Welcome to the real world!!! Although refinery documentation is much improved due to PSM regulations, P&IDs and/or piping isometrics are not perfect.
Bottom line was that the connection between the two sewers was severed and we traced the source of the leak to some coolers using once through sea water which discharged into the uncontaminated sewer. We used another ancient device to determine the exchanger leak – a bubbler. The exchanger was isolated, blanked off, heads dropped and leaky tubes plugged.
So now the youngster has some oil on his boots, some dirt on his hard hat and some oily water on his gloves. This was his first time “around the block.” He has come a long way to becoming a real Process Engineer!