I've been in the electrical equipment game for 8 years now. I handle transformer and protection relay orders for industrial facilities. In my first year (2017), I thought I had it figured out. I thought the key to success was finding the cheapest part that would 'probably work.' I was wrong. Three blown control panels and a $15,000 emergency replacement later, I learned a hard truth: in emergency situations, paying for delivery certainty is the only smart decision.
The Problem You Think You Have: Picking a Surge Protector
So you're looking at a coax lightning surge protector. Or maybe you're wondering, "What is a surge protector, really?" You've probably got a list of specs, a budget number in your head, and a deadline breathing down your neck.
I get it. I've been there. Last September, I needed a replacement for a blown GE potential transformer at a substation. The meter was dead, and the operators were screaming for a fix. I had two options: a standard unit from a distributor with an 8-week lead time, or a premium GE transformer with monitoring capabilities that could ship in 5 days. The premium unit cost nearly double. Easy choice, right? Pick the cheaper one.
That's exactly what I did. And it's exactly where I went wrong.
The Deeper Issue: What Nobody Tells You About "Protection"
Here's the part that took me years to understand. The problem isn't just picking the wrong device from a catalog. It's about the entire protection ecosystem.
1. The Surge Protector Isn't the Problem
When people search for "what is a surge protector," they think it's a simple piece of hardware that blocks spikes. In reality, a coax lightning surge protector at a cell tower or industrial site is only as good as the grounding system and the equipment it's feeding. I once saw a $12 surge protector fail because it was connected to a poorly bonded ground. The lightning still jumped the gap. The protector didn't work because the environment wasn't designed for it.
2. The Transformer Match Is Everything
A GE potential transformer (PT) is used to step down high voltage to a safe level for meters and relays. But here's the kicker: if you mismatch the VA rating or the accuracy class, your relay will give you garbage readings. You'll think you have a fault. You'll trigger a trip. You'll shut down a line for no reason. The conventional wisdom says to look at the voltage ratio. My experience says look at the burden—the load the PT is driving. If it's overloaded, everything downstream is blind.
3. The Relay Programming Is a Black Hole
GE Multilin transformer protection relay models (like the 850, 845, and 750) are incredibly powerful. But I've seen techs install them and never touch the settings. They leave the default parameters in place. The relay is effectively useless. One facility near Houston had a Multilin 850 set to the factory defaults for three years. A minor overcurrent fault that should have been isolated became a cascading failure. The plant was down for 12 hours.
So the real problem isn't finding a surge protector or a transformer. The real problem is that the industry doesn't talk enough about integration, matching, and commissioning. We buy parts like they're standalone. They're not.
The Cost of Getting It Wrong
Let me give you a concrete example from my own notebook (note to self: I really should digitize these).
In March 2024, our team ordered a replacement GE distribution transformer for a hospital emergency backup system. It was a 500-kVA unit with a standard lead time of 10 weeks. The manufacturer offered a rush option at a 25% premium—ship in 5 weeks. The purchasing agent balked at the extra $4,500. He said, "We can wait. The budget is tight."
We waited 8 weeks. Then we got a call: the transformer was delayed another 6 weeks due to a material shortage. The hospital's existing unit failed during week 9. They had to run on the backup generator for 8 days straight—at a fuel cost of about $6,000 per day. The generator's maintenance schedule got wrecked. The hospital almost had to cancel elective surgeries. The $4,500 we saved? It cost the client nearly $70,000 in direct expenses, plus the reputational damage.
That's when I stopped arguing about rush fees. The rush fee buys certainty. And certainty is priceless when the lights go out.
Another time, I specified a non-GE current transformer for a feeder monitoring application. It was 40% cheaper than the equivalent GE current transformer from the Multilin catalog. The first unit drifted out of spec within six months. The second unit had a cracked core after a fault. The third unit was fine, but we had wasted $2,100 in installation labor and taken a production line offline twice. The GE unit, which cost $900 more upfront? Zero issues in three years. Expensive is cheap in hindsight.
The Short Solution: How to Buy Protection (and Transformers) Right
If you have made it this far, you already know what to do. But let me spell it out so we both feel better.
Rule 1: When the deadline is real, pay for the premium delivery.
If a missed delivery will cost more than 1.5x the rush fee, pay the fee. Every time. The math doesn't lie.
Rule 2: Match the entire chain, not just the part number.
If you are buying a GE potential transformer to feed a Multilin 850 relay, verify the burden, accuracy class, and thermal rating. Do not assume the catalog match is the system match.
Rule 3: Commission the relay properly.
The best GE Multilin transformer protection relay model is worthless if it is set to factory defaults. Program it for your application or hire someone who can.
Rule 4: For surge protectors, fix the ground first.
A coax lightning surge protector is the last line of defense, not the first. If the ground path is bad, the protector will fail. Check the bond before you buy the part.
I have a checklist now (47 items, updated monthly). I wish I had it in 2017. But that is the nature of this business—you learn, you document, and you try not to repeat the expensive lessons.
Leave a Reply