What do Oil and Gas Risk Filings Reveal about Maintenance?
Martin Holm Nielsen, Cofounder and CEO of Arkyn (arkyn.io)
Posted 7/16/2026
All listed oil and gas companies are required by law to tell investors any factors that could potentially hurt the business. Lawyers review every word, then executives sign off on the result. Risk-factor disclosures are some of the most carefully considered texts a company produces.
Additionally, they are useful reading for maintenance and reliability professionals. If you want a real voice in budget and priority discussions, you have to connect your objectives to the risks the business already cares about. The risk filings tell you, in your leadership’s own approved language, exactly what they are.
Earlier this year my team read the 2025 risk disclosures of thirteen of the largest companies in oil and gas: ExxonMobil, Chevron, ConocoPhillips, Halliburton, Petrobras, BP, TotalEnergies, Equinor, Eni, SLB, Occidental, Marathon Petroleum, and Shell.
Six themes appeared in every single filing:
- price and margin volatility
- regulation and compliance
- climate and the energy transition
- operational hazards and safety
- cybersecurity
- equipment failure or operational disruption
The companies rarely name maintenance directly, but if you read the filings through a maintenance lens, it is everywhere. Here is what the disclosures reveal, and what maintenance and reliability teams can do with it.
Take Equipment Failure Risk Seriously
All thirteen filings name equipment failure or operational disruption as a material risk. Marathon Petroleum covers the full range in one sentence:
“Our operations are subject to business interruptions, such as scheduled and unscheduled refinery turnarounds, unplanned maintenance, explosions, fires, refinery or pipeline releases, product quality incidents, power outages, severe weather, labor disputes, acts of terrorism, or other natural or man-made disasters.” (Marathon Petroleum, FY2025 Form 10-K, Item 1A)
The cost is easy to underestimate until you put a number on it.
Take an illustrative refinery losing 250,000 barrels of throughput to a single day of unplanned downtime. At a gross refining margin of $10 per barrel, that is roughly $2.5 million of lost gross margin in one day before restart costs, overtime, and contractual penalties.
The instinct under that kind of pressure is to increase overhaul frequency, but the evidence says otherwise. Nowlan and Heap’s 1978 study for the US Department of Defense, the report that established reliability-centered maintenance, showed that most failure modes are not age related. Fixed-interval overhauls cannot prevent those failures, and intrusive maintenance can actually introduce new ones.
What does work is acting inside the P-F interval: the window between the moment a problem becomes detectable through a vibration signature, a temperature rise, or a pressure change, and the moment the asset functionally fails.
Condition-based and predictive approaches use that window to concentrate effort on the equipment that is genuinely degrading and on the failures that carry the most safety, environmental, and production risk.
Most maintenance teams I meet are stuck in a blend of reactive firefighting and fixed-interval preventive work. The execution data needed to improve maintenance maturity is simply not being captured, because reporting from the field is too slow and too painful.
Turnarounds Move Quarterly Results
Refiners and integrated operators flag planned maintenance events as material risks, and for good reason: turnarounds are high-cost, high-stakes, and visible to investors when they slip. McKinsey has estimated that better shutdown and turnaround management can deliver schedule and cost improvements of up to 30 percent (The upside of downtime – McKinsey).
Turnarounds rarely slip because the original plan was bad. They slip because the plan loses the plot: scope creeps, priorities shift, and by week two nobody is working from the same version of the truth.
Compliance is an Execution Problem
Every company in the analysis flags regulatory complexity as material. ConocoPhillips states the maintenance connection plainly:
“We have incurred and will continue to incur substantial capital, operating and maintenance, and remediation expenditures as a result of these laws and regulations.” (ConocoPhillips, FY2025 Form 10-K, Item 1A)
For maintenance teams, the exposure is concrete. Compliance requires inspections that happen on schedule, findings that get tracked to closure, and an audit trail that proves both.
If your integrity management runs on spreadsheets, email, and paper, audits are a lot harder to handle than if they were digitized.
The Talent Blind Spot
Here is the finding that surprised us. Only seven of the 13 companies analyzed name talent, skills, or workforce capability as a distinct risk: Halliburton, SLB, BP, Equinor, Petrobras, Shell, and TotalEnergies. Several mention people only in passing inside other risks.
For a sector whose frontline workforce is aging and retiring, that gap between the filings and reality is striking. When your most experienced instrument technician retires, the labor hours are replaceable. The decades of pattern recognition, troubleshooting instinct, and site-specific knowledge are not, unless they have been captured in procedures, checklists, and job plans while that person was still on the payroll.
Where companies do name the risk, the practical symptoms they describe are the ones every maintenance manager recognizes: slower work completion, higher rework, and more variable execution. Knowledge capture works best when it is a by-product of doing the job, through digital work instructions and checklists that record as-found and as-left detail as part of completing the work, rather than an admin task saved for the end of the shift.
Bad Data Underwrites All of it
One risk hides underneath the others: data your teams cannot trust. Incorrect equipment hierarchies, functional locations that do not match reality, bill-of-materials errors, and phantom inventory where the system says the part is on the shelf and the shelf disagrees. Gartner has put the cost of poor data quality at a minimum of $12.9 million per year for the average organization (Gartner, Data Quality).
There is a usability angle, too. Peer-reviewed research in the Journal of Quality in Maintenance Engineering found that hard-to-use CMMS interfaces promote user error and leave people unwilling to use the system regularly (Tretten and Karim, 2014). When technicians stop trusting or using the system, execution history stops being captured, and every data-dependent ambition further up the maturity curve, from condition-based maintenance to AI, is built on sand.
How to Use this at Your Site.
Two suggestions for turning this analysis into action.
First, build a risk-based maintenance approach. Prioritize the work that reduces enterprise risk across safety, environmental exposure, production impact, and compliance, and describe that work in the same terms your leadership used in the annual report. A backlog item framed as “mitigates the operational disruption risk disclosed in Item 1A” gets a different hearing than one framed as a maintenance cost.
Second, connect your frontline in both directions. Push complete work context out to technicians so jobs go faster and capture clean execution data back so the record matches reality. Every theme above, from reliability to compliance to knowledge retention, depends on that loop.
If you haven’t, read your own company’s risk factors this week. It takes an hour, and it hands you the business case for maintenance that your executives have already signed.
The full analysis, including filing excerpts for each theme and a maintenance maturity model, is available at arkyn.io/blog/oil-and-gas-risk-report-2026.
Related Articles
Developing Marketable Engineering Skills
Eliminating Conveyor Concerns
Installations and Inspections of Corner-grounded Systems
Understanding Shaft Alignment: Thermal Growth
Revitalizing an Aging Grounding System
The Trouble with Torque in Electrical Connections
