After 13 years, I can divide my P&ID mistakes into two categories: the ones I caught before construction, and the ones construction caught for me.
The second category is more expensive.
1. The Check Valve That Wasn’t
I once drew a check valve backward on a pump discharge line. Noticed it during a review — no, I didn’t. The pipe fitter noticed it during installation. He called the site engineer, the site engineer called me, and the conversation was basically: “You sure about this?”
I wasn’t. He flipped it around.
Cost: a 10-minute phone call and some embarrassment. If he hadn’t caught it and the pump had been started against a closed check valve, the pump casing would have overpressured. That’s a $12,000 repair and two weeks of downtime.
Lesson: check valves are the easiest thing to get wrong on a P&ID because they’re a tiny symbol with a tiny arrow. Now I trace every check valve with my finger and verify that the arrow points toward the destination, not the source. Sounds stupid. Works.
2. The Missing Drain
Between two block valves on a strainer — the isolation valves that let you clean the strainer without draining the entire system — I forgot to put a drain valve between them.
The operators discovered this the first time they isolated the strainer for cleaning. They closed both block valves, opened the strainer cover, and got a face full of hot condensate that had been trapped between the two closed valves with no way to drain.
Nobody was hurt. But I got a very direct phone call from a very unhappy shift supervisor. The drain valve cost about $80. The rework to hot-tap one into an existing pipe cost about $4,000. The loss of operator trust: you can’t put a price on that.
Now I have a rule: between any two block valves, there’s either a drain or a very good reason why not. No exceptions.
3. The Control Valve With No Bypass
I once put a 6-inch control valve in a cooling water line with no block-and-bypass arrangement. The logic was: “cooling water isn’t critical, we can shut down the whole loop for maintenance.”
First summer after startup, the control valve trim eroded — which cooling water valves do, because cooling water is never as clean as the design assumes. The valve wouldn’t fully close. We couldn’t isolate it without shutting down the entire cooling water loop. Which meant shutting down the entire production line. In August. For a $1,500 valve trim replacement that should have taken 4 hours.
The downtime cost about $180,000 in lost production. The block valves and bypass would have cost $3,000 during construction. This is not a complicated calculation, and I got it wrong because I was optimizing for the wrong thing.
Every control valve in continuous service gets a block-and-bypass. No, I don’t care if it’s “just water.” Yes, even that one.
4. The Instrument That Was Invisible
I drew a pressure transmitter on a reactor discharge line. Correctly tagged, correctly ranged, correctly located on the P&ID. What I didn’t check: the actual physical location.
When the plant was built, that transmitter ended up behind a structural column, 4 meters above the nearest platform, with the display facing away from the access ladder. To read it, an operator had to climb the ladder, lean around the column, and squint at the gauge upside down.
The operators didn’t bother. They ran that reactor based on a pressure gauge 20 meters downstream — which read 0.3 bar lower due to line losses. The reactor ran at a slightly higher pressure than intended for two years before anyone noticed. The higher pressure reduced catalyst life by maybe 15%, costing something like $40,000 in extra catalyst changes.
An instrument that can’t be read doesn’t exist. Now when I review P&IDs, I ask: “Show me how an operator reaches this instrument. Show me a picture of the actual installation.” If nobody can produce that picture, the instrument location isn’t designed yet — it’s just a bubble on paper.
These four mistakes have something in common: none of them were caused by not knowing the theory. I knew how check valves worked. I knew strainers needed drains. I knew control valves needed bypasses. I knew instruments needed to be accessible.
The theory wasn’t the problem. The problem was that theory lives in textbooks, and mistakes live in the gap between what you know and what you actually draw on the page at 4:30 PM on a Thursday when you’re trying to finish a P&ID before the review meeting.
The fix isn’t more knowledge. It’s a checklist, a second pair of eyes, and the humility to assume you’ve made these mistakes even when you’re sure you haven’t.