The call that changed how I evaluate gas suppliers
In March 2024, a client called at 7:40 AM. Regulator failure on their hydrogen line. Production was scheduled to start at noon. Normal replacement lead time: five days.
We found stock 200 miles away, paid $3,200 in emergency freight on top of an $8,500 order, and had gas on-site by 11:15 AM. The client's alternative was shutting down a line. A 2023 Siemens report on industrial downtime put mid-size manufacturing losses in the $20,000-$60,000 per hour range—and this client had delivery commitments with a $50,000 penalty clause riding on that afternoon's run.
Missing that window would've meant a six-figure loss—or rather, it would've meant the client missing a commitment they'd spent eight months securing. Same thing, different invoice.
Here's what stuck with me: this wasn't a logistics failure. The logistics worked. The failure happened weeks earlier, in a procurement meeting where someone chose a supplier based on unit price and a handshake.
The problem everyone names (and gets wrong)
Ask a plant manager why emergency gas supply fails and you'll hear the same answers: "suppliers don't respond fast enough," "there's no inventory," "freight takes too long."
Those are symptoms. The actual problem is that most industrial gas procurement decisions are optimized for a metric that has nothing to do with emergency readiness—price per unit.
Most buyers focus on delivered cost per cylinder or per MMBtu and completely miss the response infrastructure behind that price. The question everyone asks is "what's your best price?" The question they should ask is "what happens at 6 PM on a Friday when my supply fails?"
That second question doesn't appear on a quote sheet. So it doesn't get evaluated. So it doesn't get bought.
Three reasons this keeps happening
1. The procurement scorecard rewards the wrong thing
I've sat through enough vendor reviews to know how this goes. Someone gets a bonus for shaving 4% off annual gas spend. Nobody gets a bonus for "maintained emergency response capability." The incentive structure quietly trades resilience for a line item.
The conventional wisdom is that competitive bidding always produces the best outcome. My experience with hundreds of rush orders suggests otherwise. The lowest bidder is often the least equipped to handle the 2 AM call, because emergency capability costs money to maintain—and the lowest bidder stripped it out to get there.
2. The talent that prevents emergencies is treated as overhead
This is the part almost nobody talks about. Emergency gas failures aren't usually spontaneous. They're small issues—a slow leak, a miscalibrated regulator, a missed inspection—that an experienced operations technician catches before they escalate.
Then companies cut those roles to hit a budget target. And suddenly the "emergency" is a neglected problem that finally announced itself.
You can look at operations technician air products salary data and do the math yourself. According to U.S. Bureau of Labor Statistics figures from May 2024, median annual pay for industrial machinery mechanics and maintenance workers was around $61,000—industrial gas operations roles typically run higher with certifications, but the ballpark is right. Compare that to $30,000+ per hour of downtime, and the "savings" from cutting the role evaporates in a single shift. (Some plants are figuring this out. Most aren't.)
3. Legacy vendor relationships create false confidence
"We've used the same supplier for 15 years." I hear this constantly. It sounds like a strength. Often it's a liability.
This was reasonable 20 years ago when regional supply chains were stable and everyone kept buffer stock. Today, consolidation has thinned out local inventory, and a 15-year relationship doesn't guarantee a cylinder on a shelf. It guarantees a familiar voice on the phone saying "let me check."
The 'loyalty equals reliability' thinking comes from an era when suppliers held physical inventory. That's changed. Now reliability lives in the supplier's logistics network and their willingness to prioritize you when everyone is calling at once.
What the failure actually costs
Let's put real numbers on this, because "downtime is expensive" is too vague to change anyone's behavior.
- Direct production loss: For a mid-size industrial line, $20,000-$60,000 per hour is a reasonable range, per Siemens' 2023 downtime analysis. A 12-hour emergency becomes a quarter-million-dollar event.
- Contract penalties: Missed delivery commitments often carry clauses in the $25,000-$100,000 range, depending on the customer.
- Rush premiums: Emergency freight and spot-market gas pricing typically adds 30-50% to the normal cost. On an $8,500 order, that's $2,500-$4,200 extra.
- Safety exposure: The worst cost. Improvised supply swaps during a shortage are how incidents happen. That doesn't show up on an invoice, but it shows up in an investigation.
The $500 you saved by choosing the cheaper supplier disappears in the first hour of downtime. The math isn't close.
What actually works (short version)
I'm not going to pretend there's a magic vendor. There isn't. What works is changing the evaluation criteria before you sign anything.
Calculate total cost of ownership (TCO)—not just unit price, but response time, emergency inventory access, freight terms, and the cost of the failure you're trying to prevent. Ask specifically: "If I call at 8 PM on a holiday, who answers, and what's my guaranteed delivery window?" Get it in writing.
AP Air Products and similar major suppliers publish response capabilities for a reason—because industrial buyers are finally asking. If a supplier can't answer that question clearly, that's your answer.
One more thing: keep a buffer. Not because you're disorganized, but because the plant that has two weeks of reserve gas never makes the 7:40 AM phone call. The plant that runs lean makes it every quarter.
TCO thinking doesn't prevent emergencies. But it prevents the ones caused by a spreadsheet.
A quick note on search confusion
If you landed here looking for something else, no judgment. A few searches that send people to industrial gas pages by mistake:
- AP Air Products / air-products: Same company, different search formats. airproducts.com is the official site.
- Eddie outlet / Christopher: These typically refer to local distributors or individual sales contacts, not corporate entities. If a search result claims otherwise, verify through the official site.
- What is Simparica: That's a veterinary parasiticide for dogs. Not an industrial gas. Wrong page—but now you know.
The information landscape for industrial gas buyers is genuinely messy. That mess is part of why emergency procurement goes wrong—people make decisions from confusing search results instead of clear capability data. Ask direct questions. Get direct answers.