Engineering in the oil & gas industries is serious stuff (excuse the generalization and
motherhood). The modern design and construction of new facilities are expensive, running anywhere from 2x to
4x over budget, taking months and years over-schedule and sometimes, don't even operate successfully. The
trend was set during modern wars (i.e., mid-19th century onwards) when governments
wrote blank cheques to engineering & construction (E&C) contractors. In earlier times, the King or Pharaoh
would not pay or beheaded the contractor who consistently overran budget and schedule.
Today, of course, it's different but not necessarily better. And coming from the top, the sentiment reaches down into
the ranks of the organization. Some CEO's have even lectured their employees to not worry about cost or
schedule. The impact of that attitude burrows into the subconscience and soon tasks take a little longer, at first,
and then more as complacency sets in. Sometimes, it's explicit, sometimes subtle.
Here we give a sampling of situations that exemplify this devolution of competency, responsibility and caring.
Design-to-Fail Engineering
The design of equipment in the oil & gas industries is superbly defined in many industry standard practices
such as the ASME Codes, API, TEMA, ASTM and other Standards whether of a basic or advanced level. However,
some OEM suppliers use the opportunity to use simplified methodologies with their explicit deficiencies when advanced
methodologies are required.
Two situations in the industry are common; the use of the simplified wall thickenss design
(API 530 "thin wall") of heater pigtails which then creep fail per that design methodology at 2 to 3 year intervals and, then, requiring
multi-million dollar replacement. A design life of 11 years can typically be designed and achieved by more exacting methods.
In delayed-coke-unit ((DCU) coke drum equipment design, the skirt-to-drum attachment is susceptible to thermo-mechanical
fatigue failure (1,000 to 10,000 cycles / 3 to 30 years); yet, designers appear to specify and construct demonstrably weak joint details
which fail 2x to 5x sooner than the more robust available designs (API TR 934G).
Wacky Project Management
Real Time Performance Measurement To measure progress & performance on small projects, the
local office developed a bespoke automated reporting system that could take the automated time sheet, track
hours against the work breakdown structure (WBS) and report earned and actual performance. Projects could be
tracked daily since timesheets were completed daily and matched to budget hours. It caught the attention
of head office who, in short order, killed it in favour of the traditional system which was always out-of-date
by at least two weeks and took a separate department to administer. Needless to say, progress and performance
suffered.
Mandating Budget Overruns A rather shocking instruction involved the Engineering Manager instructing
a project engineer that success was mandated by overrunning the Client's budget by 25%. If that could not be achieved, the
project engineer had no business being in that position. By that yardstick, we have experienced many "successful"
projects, especially in certain Canadian provinces.
Wacky Engineering
The following are specific examples where the engineering is unskilled and discipline manhours are
wasted to the detriment of the facility owner but to the benefit of the contractor's bottom line.
Deficient Vessel Registration / Unskilled Engineering Practice
The widespread use of design software is allowing many to take up pressure vessel engineering. An astonishing recent incident
revealed that PSA vessels were to be designed for cyclic thermal loading. Unfortunately, the designer failed to grasp
the concept and used the difference in operating & design pressures to define a "cyclic" load. What's even more astonishing
is that the fabricator, the owner and the registration authority failed to comprehend this deficiency. In addition,
the User Design Specification was not developed for this ASME VIII Division 2 vessel ! Needless to say, there are likely many
similar vessels operating in plants across North America+.
FEA / Numerically Based Pressure Equipment Design & Analysis
FEA / numerical solution techniques have been automated to the point "design drafting" personnel are performing FEA functions;
typically with minimal data entry and a few button clicks. Lengthy "evaluation" reports can be autogenerated with
another few clicks which will contain a number of full-coloured graphics. The prior example above is one such case, another
involved designing coker drum skirt reinforcement which failed to identify potential fatigue failure from both the
original design concept and later, an after-market design modification which exacerbated the original flawed philosophy.
In both instances, fundamental design principles were overlooked.
Piping Stress Analysis
Piping stress is a nice money-maker. The EC typically has a "piping analysis criteria" document which defines the
piping lines that "require stress analysis". Careful review and evaluation will reveal that almost every piping line
requires formal stress analysis. When the analysis is performed, anywhere from 40 to 80 hours of manpower are allotted
for a line, just for the "analyst". Experience and Piping Code alternatives to this formal review and not well
received. A piping line, if analysis is essential, should take about 24 hours to complete. Many lines can be assessed
by an experienced piping designer or comparison to existing systems.
Piping Stress Analysis
Piping stress is a nice money-maker. The EC typically has a "piping analysis criteria" document which defines the
piping lines that "require stress analysis". Careful review and evaluation will reveal that almost every piping line
requires formal stress analysis. When the analysis is performed, anywhere from 40 to 80 hours of manpower are allotted
for a line, just for the "analyst". Experience and Piping Code alternatives to this formal review and not well
received. A piping line, if analysis is essential, should take about 24 hours to complete. Many lines can be assessed
by an experienced piping designer or comparison to existing systems.
Underground Piping Stress Analysis in Terminal Tank Farm
The task was to run a slightly elevated temperature line from one large-diament tank to another in the
pipeline operator's tank farm. The analyst used a well-known software aided engineering tool.
After some six-months of effort, the line was still not approved. The root problem was that the line was
underground and large restraint stiffnesses were being generated by default calculation. Of course, unknown to the analyst,
almost any routing was acceptable using the closed-form calculation methods of ASME B31.8 or CSA Z-662.
The piping software was overly conservative and was actually a misuse / abuse of the software.
Underground Piping Transient Hydraulic and Stress Analysis
A similar task to the previous example but a transient hydraulic analysis was required. For some unknown reason, it took
several teams almost a year to engineer an acceptable routing for the NPS 12 crude oil line. The cost was $1 million!
Utility Boiler Heat Loss Calculation
The client required a heat loss calculation for sizing of ventilation vents for the boiler house. Incidental was to calculate
a heat loss and efficiency rating for the industrial power boiler. An analyst was assigned and a year later a value was determined;
the analyst had binders full of heat transfer calculations for various surfaces comprising the boiler which was 8 or so stories high.
The engineering cost was north of $250,000. A check of the calculations was requested at which time a more experienced engineer
used the ASME PTC documents, knowledge of industry association standards and determined a value within 2 days that was within 5%
of the "long method" calculation. The "industry" method is based on thousands of units had proven itself over 10 decades of use.
OTSG Rerate
The situation required rerating of an OTSG to higher pressure. The head office engineer decided to use the equipment "designer".
After a year, the task was still incomplete and the execution strategy very much foundering and clear as an Alberta snowstorm.
It was clear, the OEM was stringing the owner along and "milking" it. The "old hand" engineer, who had design experience
with power boilers, was finally turned loose. Within 3 months, the unit was rerated and approved by the regulator. Later, when
there was interest in re-rating the HRSG, the owner turned to the "old hand" again. The rerating opportunity was
confirmed and production was increased by 2-½% within 4 months. The calculations were developed from 1st principles
as software was not available. Tube fin temperatures figured in limiting flue gas temperatures.
Materials Selection for DCU Coker Drums
This hydrocarbon processing technology has been in use since the late 1920's and used carbon steel drums. This was upgraded to
SA-204 C (C - ½ Mo) in the 1950's. A vexing problem was cracking of the drum and skirt and a unit quench factor (UQF) was developed
to correlate water fill rates to bulging and cracking damage. As the problem persisted, focus on the materials of construction and
various damage mechanisms was given; drum "vasing" and "coke crushing" were two mechanisms leading to increased material alloy specification.
Materials graduated to SA-387 12 (1 Cr - ½ Mo), SA-387 11 (1¼ Cr - ½ Mo-Si), SA-387 22 (2¼ Cr - 1Mo) and, then
SA-387 21 (3 Cr - 1 Mo). The engineering consultant was able, somehow, to obtain a patent for the use of various high yield strength
material; you say what?
The owner of the drums with the 3Cr really dislikes those drums - check the welding procedure!. A joint industry research program
eventually demonstrated that SA-204 C and, an alternate, SA-302 C had the appropriate material properties to
resist the high strain range cycling imposed by the water-fill operation.
Water-fill induces large elastic / plastic strains occasioned by thermal-mechanical cycling on account of 900 F drums
being repeatedly injected with water at 250 F; makes sense and intuitive UQF has merit. Work out the induced strain range for yourself.
We have more "wonderful" examples and we will post as time permits.
If you have an unusual situation which just doesn't make sense; gives us a call. Perhaps we can help, perhaps
we are just interested in hearing about another wacky engineering example in this industry.
John Aumuller, P. Eng., Ph. D.