The refinery's documentation was organized and generally pretty good but its isometrics were terrible. Over the next few months our efforts to upgrade its desalter were stalled several times because of incomplete, conflicting and often dead-wrong isometrics.[pullquote]
Isometric drawings can be the most valuable construction aid available besides Process and Instrumentation Diagrams (P&IDs). Stick figures and familiar symbols can easily represent construction details sufficient for a welder to build a process, a contractor to bid a project, and an engineer to evaluate process feasibility.
Unfortunately, some people consider isometrics an anachronism. During a hydrotreater expansion the overall contractor was convinced 3D modeling would solve all problems with our layout. I was amused. The rest of the team were not when I asked if 3D modeling would be available to the welders actually building the plant. The isometrics when reviewed highlighted numerous support issues overlooked by the modelers.
However, isometrics won't solve any problems if done poorly. Bad isometric drawings could cost your company a fortune. Here's an example from one hydraulic study of a 2,500-ft brine line we completed.
Our analysis should have required about two weeks. Instead, it took about six weeks and hundreds of man-hours. Why? Instead of pipe lengths, I had to work with distances between columns; a convenience if you're building but not if you want to analyze later. Existing valves were marked as demo-ed. Elevations were given as vague measurements between objects; a specific datum such as distance above sea level would have been useful. Over 400 feet of the pipeline was based on a plan view because no isometric drawings existed. The analysis involved three mechanical engineers for three weeks and a chemical engineer for the models for six weeks. In addition, department oversight and quality control took, maybe, two weeks. It probably wound up costing around $60,000. In comparison, with good isometrics, the job would require two mechanical engineers and a chemical engineer for a week and a week of quality control — which would run about $20,000. Worse yet, the refinery will have to repeat this experience for every future project.
So, now that you understand what's at stake, consider ten steps that can improve isometrics:
1) Identify each run of pipe in lengths of a common unit. Avoid mixing inches and feet — for small sections use inches.
2) Include equipment lengths, excluding pipe flanges and gaskets, separately.
3) Ban the term "field verification" — it leads to sloppy design. Remember, if you can't design for it you can't build from it.
4) Allow some slop for construction and threading.
5) Keep demo drawings separate.
6) Add supports with reference to specific, uniform details, such as "dummy support for carbon steel pipe."
7) If you're not sure how a pipe is connected and it's not part of your immediate concern, don't guess — leave it blank on the drawing with a note to describe the problem.
8) Coordinate the isometrics with plan and elevation drawings. Ideally, you should have a plan view grid of individual isometric drawings.
9) Take advantage of new technology by integrating, rather than replacing, the old. Use 3D modeling to spot obstructions.
10) Learn by doing. Try setting up a hydraulic model of the piping using the isometric — chances are you'll find missing details you'll want to add pipe drawings.
Also, consider including basic drawing symbols for pipe assembly: two vertical parallel lines for flanged pipe; an enlarged dot for welded pipe; and a single vertical line for a threaded connection. You might have to get more specialized for construction that involves butt-welding, socket-welded pipe, NPT or DIN threading, and a variety of flanges (slip-on, raised-face, weld-neck, lap, threaded, etc.).
Don't forget to specify the gaskets! Using three 3/8-in. gaskets instead of three ¼-in. ones adds 3/8 in. to the overall pipe length. This might not sound like much but it affects thermal expansion and even constructability if the piping is tight. Be specific.
These details are best handled in a Bill of Materials table; make sure the measurements on the drawing are true to the actual fittings. This is one difference between P&IDs and isometrics. You can show a 6X2 reducer on a P&ID, but on an isometric drawing no such thing exists — a 6X2 reducer becomes a pair of 6X3, 3X2 reducers or 6X4, 4X2 reducers. More detail is required.
Being involved in the piping details will improve your chances of project success.
DIRK WILLARD is a Chemical Processing contributing editor. You can e-mail him at [email protected].