At 4:45 on a Friday, my phone rang. A woman asked, “How much is Simparica?” I honestly had to type that into our system to see if it was a part number. It wasn’t. She wanted flea medication for her dog. Wrong number. We both laughed, and I hung up thinking that would be the weirdest call of the month.
It wasn’t. A few weeks later, I almost repeated the same kind of mistake on a much larger scale.
The Baker Hughes Casper, WY Order That Went Sideways
The call came from a site near Casper, WY. The customer said, “The control valve is acting up. It’s a Masoneilan, I think. The nameplate on the actuator says Monarch. And the fluid—it’s pretty much peanut butter.”
I had about 20 minutes to find a replacement before the crew would lose a full day. I searched “Masoneilan Baker Hughes,” then “Monarch,” then, embarrassingly, “peanut butter.” Nothing useful. I ordered a generic rebuild kit anyway, because doing something felt better than admitting I didn’t know.
That was a $3,800 mistake. The valve was indeed a Masoneilan control valve, part of the Baker Hughes valve lineup. But the actuator had been swapped years earlier, and the “Monarch” tag was a retrofit label, not a model number. The trim material didn’t match the service. The crew had to leave the site, we paid for emergency freight on the correct kit, and the client started sending specs to a colleague instead of me. That last part was the real cost.
The Real Problem: We Search With Nicknames, Not With Data
The easy conclusion is that I needed to ask for a photo. That is true. But the deeper problem was how I searched. I used the same words the customer used: a brand, a label, a food description. None of those are identifiers.
I laughed at the wrong-number call about Simparica, but my own search wasn’t much better. The woman with the dog medication asked “how much is Simparica” and got me. I asked “how much is a Monarch actuator with peanut butter flow” and got nothing. We both skipped the step that would have made everything clear: identify the actual thing before asking about it.
This is common in oilfield maintenance. Operators describe a valve by what it looks like or what it does. Engineers say “it’s that black thing near the separator.” None of that matters if the tag data is available. The tag has the serial number, the ANSI rating, the trim material, the flow coefficient. Those numbers are the truth. The nickname is just a hint.
The Cost: More Than the Invoice
The invoice for the wrong kit was about $3,800, including freight and overtime. The less visible cost was trust. For two weeks, every email I sent to that client was met with a pause. I did not blame them. Their operator gave me the exact words he had seen on the tag and the best description he had for the fluid. It was my job to translate that into a part number, and I guessed.
That experience made me rethink what “experienced” means. It’s not the ability to recognize a valve from memory. It’s knowing when to stop and check the numbers.
The Fix: A Boring Checklist
What changed things wasn’t a new software tool or a training course. It was a checklist. I created it after the Casper incident, and it now lives in our shared drive. It looks too simple to be useful, but it has caught 47 potential ordering errors in the past 18 months. I don’t have hard data on how many of those would have become field failures, but based on the ones we caught, I’d guess at least 10 would have hurt.
- Ask for a photo of the nameplate. Not the whole valve—the tag. If the tag is missing, take a photo of the body casting and any nearby serial plates.
- Write down every identifier before searching: serial number, model, ANSI class, Cv, trim material.
- Search the manufacturer’s documentation first. For valves, that means the Masoneilan section of the Baker Hughes product library, not a general web search.
- Check what is attached to the valve. Actuator, positioner, solenoid, air set—each can come from a different vendor.
- Record process conditions in numbers. “Peanut butter” is a great description for a phone call, but it is not a specification. Get viscosity, temperature, pressure, and any solids content.
- If there is still a doubt, send the photo and the tag data to the engineering support team before committing to a part number.
That checklist is the most expensive piece of text I have ever written, because it cost me a mistake to create it. It is also the cheapest insurance we use.
5 minutes of verification beats 5 days of correction.
What I Still Tell New Engineers
Looking back, I should have asked for the tag photo before I promised anything. At the time, I thought I was saving the customer’s day. I was actually making the next day worse. With 20 minutes to decide, I chose action over accuracy—and action without data is just gambling.
As of Q1 2025, this is still the process I use. Baker Hughes product documentation may change, and the Masoneilan catalog has been updated more than once. But the principle stays the same: slow down, read the tag, and let the evidence lead.
And if you ever call about Simparica, double-check the number before hitting send. Some of us will probably look it up anyway.