I coordinate completion work for a mid-sized independent operator, which means I spend a lot of time ordering cementing services, completion equipment, and well-barrier hardware. I’ve been doing this for 12 years, and I’ve personally made—and documented—14 significant mistakes that added up to roughly $2.8 million in wasted budget. I keep that number visible on my desk. It’s why I’m writing this article, and why my team now has a pre-job checklist that gets reviewed before every major completion operation.
If you work with downhole barriers, you’ve probably been where I was: a job goes wrong, and the first instinct is to blame the hardware. After my third expensive failure, I realized the hardware wasn’t the issue. The issue was how we thought about the whole system. That’s the story I want to share here, because in 2025 we can’t afford to keep making the same mistakes.
The Surface Problem: Everyone Blames the Storm Choke
In September 2022, we were completing a gas well in West Texas. The program called for a Halliburton storm choke—our operations superintendent’s shorthand for a subsurface safety valve, even though the modern tool is technically a surface-controlled subsurface safety valve, or SCSSV. A hydraulic control line holds the valve open; if surface pressure is lost, the valve closes automatically. It’s a simple, robust piece of equipment.
It failed its final barrier test.
My first reaction, and honestly the field crew’s too, was that the valve was defective. Halliburton’s completion team flew out, bench-tested the assembly, and showed it was fully functional. The root cause was a clamp on the control line. It had been installed too tight, pinched the line against the coupling, and after several pressure cycles the line developed a pinhole leak.
The frustrating part: I knew I should have verified the clamp torque during installation. I skipped it because we had installed the same system a dozen times. That was the one time it mattered.
The valve wasn’t the failure. The system around the valve was. That lesson is documented in the standards we should have been following. API RP 14B—the recommended practice for subsurface safety valve systems—doesn’t treat the valve as a standalone part. It treats the valve, lock mandrel, control line, clamps, and installation procedure as one system. We treated it as a part replacement. That’s the surface problem in a lot of well-integrity incidents: we look for the one broken component instead of asking how the system worked together.
Deeper Cause #1: Garbage In, Gospel Out
The same pattern shows up in cementing. After a 2021 cement job in South Texas came up short of the planned top of cement, the toolpusher told me, “Your simulator was wrong.” It wasn’t. The simulator was fine. The inputs were bad.
We had run the job design through Halliburton PACER—the cementing simulation tool used to model fluid placement, pump pressures, and displacement. The output predicted full coverage above the pay zone, with acceptable equivalent circulating densities. The bond log later showed the top of cement almost 800 feet lower than what the model predicted.
Why? Because the PACER simulation used the drilling-fluid density from the morning report. The night crew had lost 30 barrels of mud to a thief zone and quietly topped off with water to “keep the hole full.” The data we handed to the simulator was fiction. PACER gave us a perfect-looking answer to a wrong question.
This is the counterintuitive part: the model was usually right. People treat simulations as the final answer instead of a check on assumptions. A cement simulator doesn’t know what happened during the night tour. It doesn’t know the mud report was never corrected. It knows only what you tell it.
Since that job, my team follows a simple rule. Before any PACER run is approved for job design, the input data sheet has to be signed by the engineer who can explain where each number came from. That one rule caught more simulation errors than any software upgrade we’ve bought.
Deeper Cause #2: The White-Contract Trap
The most uncomfortable lesson didn’t involve equipment at all. It involved the service contract.
I’m going to use the term “white contract” here, and I want to be clear that it is not a legal category. It’s my name for a contract that looks perfect. Every deliverable is listed. Every rate is quoted. Every page is clear. Both parties sign it believing all ambiguity has been eliminated.
In 2020, we signed what looked like a white contract for a completion project. It specified the service company, the equipment model, even the software to be used for engineering design. What it didn’t specify was who revalidates equipment selection when well conditions change between contract signing and execution.
Two months before the job, offset-well data suggested the maximum expected shut-in pressure would be higher than originally assumed. The change was real. The change was documented in an operations email thread. And it never triggered a revalidation of the subsurface safety valve that had already been procured. The valve was rated for the original conditions, not the revised ones. Nobody caught it until the barrier test failed.
That failure cost about $1.1 million in unplanned intervention work and expedited replacement equipment. Here’s the part that still stings: the contract was technically clear. Too clear. It assigned responsibility for equipment supply, but not for engineering revalidation when well data changed. Both sides walked away believing the other side was responsible for catching it.
This is where the industry is evolving faster than our processes did. What was considered best practice in 2020—writing detailed scopes and assuming the service company will cover downstream changes—doesn’t fit how integrated well delivery works in 2025. Contracts need to assign the gray zones, not just the obvious scopes.
What These Mistakes Actually Cost
I mentioned $2.8 million earlier. Let me break it down, because the number only matters if you can see where the money went:
- November 2020: A subsurface safety valve procured for the wrong pressure envelope because a contract gap prevented revalidation. $1.1 million in interventions and expedited replacements.
- March 2021: A cement job designed with Halliburton PACER, fed inaccurate mud data by our own night crew. $860,000 in remedial cementing, rig time, and logging.
- September 2022: A control-line clamp that pinched a line and caused the storm choke to fail its barrier test. $720,000 in intervention costs plus a two-week delay.
Those three incidents alone total $2.68 million. When I add the smaller errors—the wrong spacer volume, the overlooked torque spec, the procurement mistake that had to be fixed in 24 hours—the total passes $2.8 million.
What the spreadsheet doesn’t show is credibility damage. When a barrier is compromised, the question stops being about dollars and becomes whether your team can be trusted to deliver a well safely. That’s a much harder cost to recover from.
What Fixed It (And Why It’s Boring)
There is no elegant technology fix for this. We tried better software. We tried more meetings. What actually worked was a checklist—an aggressively specific pre-job checklist.
The checklist asks questions like:
- Who provided each input to the PACER model, and can they defend it?
- Has the storm choke’s operating envelope been checked against the latest well conditions, not the conditions at contract signing?
- Who is responsible if well data changes after procurement?
- Has the snubbing jack—or any lifting equipment—been load-tested within the required period?
- Does the service contract name a single owner for system-level barrier validation?
It sounds simple because it is simple. The hard part was admitting the failures weren’t caused by our vendors; they were caused by our process gaps. Halliburton’s equipment performed as designed in every case. The storm choke didn’t fail. The PACER simulation didn’t lie. The contracts didn’t break themselves. We did.
Since we started this process in Q1 2024, the checklist has caught 47 potential errors in 18 months (as of January 2025, at least). The one that made it all worthwhile was a snubbing jack with an expired load-test certificate. A different crew had used that jack for two weeks without anyone asking to see the paperwork. One question on the checklist kept a potentially serious equipment failure from happening on our watch.
Bottom Line
If you take one thing from this article, make it this: the next well-integrity failure you deal with probably won’t be caused by the Halliburton storm choke in the well, the PACER model on your desktop, or the contract in your file cabinet. It will be caused by the gap between those things—the space where data, accountability, and assumptions overlap.
I wrote this because I’d rather you learn from my $2.8 million mistake than from your own. The fundamentals haven’t changed: good data, clear accountability, and well barriers treated as systems, not parts.