Look, I'll be straight with you. When a Siemens PLC fails on the line, the first question you hear isn't "what broke?" It's "how fast can we fix it, and what's it going to cost?"
I'm a procurement manager at a 60-person industrial automation company. I manage our PLC and controls hardware budget—roughly $120,000 a year—and I've been tracking every order, every repair, and every service contract for the past 6 years. That's over $180,000 in cumulative spending across 15+ vendors.
I didn't build a TCO (total cost of ownership) spreadsheet because I'm a spreadsheet nerd. I built it because I got burned by repair quotes that didn't include everything. After enough surprises, you start asking better questions. This article is basically a summary of those questions—and the answers.
I'm comparing three paths when a Siemens PLC or drive fails:
I'll compare them along four dimensions: upfront cost, downtime and engineering labor, reliability risk, and hidden fees. Price alone never tells the whole story.
Here's the thing: repair quotes usually run between 40% and 70% of the price of a new unit. From the actual quotes I've collected:
On paper, repair wins. And for legacy S7-300 models that Siemens no longer sells, repair or refurbished units are often the only road—a new S7-315 just isn't available.
But watch the fine print. In Q2 2024, I gathered 11 quotes from 6 repair vendors for three different failures. Seven of them added a bench test fee ($60–$120), three added rush handling ($50–$150), and two charged extra for return shipping insurance. One vendor's initial quote was 38% of the new-unit price. After all add-ons, it was 61%.
The vendor who lists all fees upfront—even if the total looks higher—usually costs less in the end.
Mini-conclusion: Repair wins on sticker price, but only if you ask "what's NOT included?" before you agree. Get an itemized quote in writing.
The real cost of a repair isn't just the invoice. It's the downtime while the unit is away, the labor to reinstall it, and the TIA Portal work to verify or reconfigure the program.
A send-out repair typically takes 7–14 days door-to-door. If your line is down, that's one to two weeks of either downtime or running on a temporary replacement. A new S7-1200 or S7-1500 from a stocking distributor, on the other hand, can arrive in 1–3 days.
On-site repair is faster—a good technician can often fix a power supply or I/O fault in 4–8 hours. But the day rate is steep. From quotes I've seen, on-site service runs $800–$1,500 per day plus travel, plus parts. Worth it? Only if your downtime cost is high.
Then there's engineering time. After any repair or replacement, someone has to verify the program, check the hardware configuration, and run a test cycle. Loaded labor for a controls engineer runs $85–$120/hour. A 4-hour verification adds $340–$480 to either path.
Mini-conclusion: If downtime costs you more than roughly $500/hour, buy speed: on-site service or new replacement. Send-out repair becomes the most expensive option when you count the waiting time.
I'm not an electrical engineer, so I can't tell you exactly why a repaired PLC fails again. What I can tell you, from six years of purchase orders and repair records, is this: 23% of the repaired units in my tracking required a second repair within 18 months. Nearly one in four.
That's a big deal, because re-failure isn't just another repair cost. It's another production stop, another service call, another round of verification. When I itemized a specific S7-300 CPU failure in March 2023:
Here's the surprise: for older S7-300 and S7-400 CPUs, that math flips. Replacing a legacy CPU usually means rewiring the panel and rewriting the program for a different platform. That's a full migration project, not a repair. In that case, repairing the existing CPU or buying a refurbished identical unit is the rational move—even with the re-failure risk.
Mini-conclusion: For active platforms (S7-1200, S7-1500), replacement beats repair in reliability-adjusted cost. For legacy platforms, repair or refurbish is the only practical option—but start planning the migration before the next failure.
Bundled service is popular, especially in Asia, where vendors offer the full cycle—purchase, sales, repair, and on-site service (구입, 판매, 수리, 출장서비스)—under one agreement. It's convenient. The question is whether it's cheaper than paying per incident.
From the contracts I've reviewed, an annual on-site service contract for a plant with 10–20 Siemens PLCs runs roughly 8–15% of installed equipment value per year. That's not outrageous, but it only pays off if you use it.
After running our own numbers, here's the threshold I use: fewer than 10 PLCs or fewer than 2 service calls per year → pay per incident. More than 20 PLCs, or a 24/7 operation → the contract is worth it, because a single major unplanned stop can cost more than the annual fee.
One more factor: if you've got a TIA Portal maintenance program in place, remote diagnostics can solve 50–60% of issues without a visit. That shifts the math even further toward pay-per-incident.
The question isn't "can we afford a service contract?" It's "what does one unplanned visit actually cost us?"
Here's the decision framework I actually use. It's simple on purpose:
Pricing above reflects quotes I received in Q2 2024 and Q1 2025. Siemens updates product lists and end-of-life dates regularly, so verify current prices and lifecycle status at support.industry.siemens.com before you commit.
At the end of the day, my job isn't to buy the cheapest PLC. It's to buy the one that costs the least over the full lifecycle. Repair vs. replace is never just about the invoice—it's about what happens after. Calculate that honestly, and the right answer gets pretty clear.