Protecting Your Investment: A GE Transformer Engineer's Guide to Avoiding Costly Protection Setup Mistakes

Why Your Transformer's Relay Setup Needs a Second Look (Even If It's a GE)

I've been handling GE transformer orders for about eight years. In my first year (2017), I made the classic mistake of thinking the GE 850 transformer relay was a 'set it and forget it' device. I assumed that because it was a GE product, the default settings would be fine for any application.

That assumption cost us roughly $3,200 and a two-week project delay. The vendor failure—well, my failure—in May 2017 changed how I think about protection setup permanently. We had a 15 MVA power transformer feeding a critical industrial facility. The GE Multilin 850 was in place, but the user-adjustable settings for the overcurrent and differential protection were left at the out-of-the-box values. A downstream fault occurred. The relay didn't trip in time because the coordinating devices weren't considered. The transformer took a hit. Not a catastrophic failure, but enough damage to require an outage and a rewind.

Since then, I've personally documented 22 major protection-related mistakes across our project portfolio. Now I maintain our team's checklist. If you're quoting or installing a GE transformer protection relay, this guide is for you.

The Problem: 'One Size Fits All' Doesn't Work for Protection Relays

There's no universal answer for how to configure your GE relay. It depends on what you're protecting, the criticality of the load, and the network environment. Below, I've broken this down into three common scenarios I see on almost every project.

Scenario A: New Installation, Critical Power Transformer (e.g., for a Hospital or Data Center)

My recommendation: Use the full suite of GE Multilin 850/845 functions, including advanced differential protection (87T), overcurrent (50/51), and overexcitation (24).

In Q3 2024, we commissioned a new 30 MVA unit for a hospital. We configured the GE 850 transformer relay with the differential setting dialed in to 0.3 per unit (verified using the transformer's nameplate). Plus, we enabled the breaker failure protection (50BF) function. Why? Because in a critical environment, a failure to clear a fault is not an option.

The key here is the commissioning test. You have to inject actual currents and verify the trip curves. I learned this the hard way: in 2020, I skipped the primary injection test on a distribution transformer because I was behind schedule. I figured, 'what are the odds?' The odds caught up with me when the relay miscoordinated with a downstream fuse during a test. That error cost $890 in redo plus a 1-week delay. (Mental note: never skip the primary injection.)

Scenario B: Retrofitting Protection on an Existing Transformer (e.g., Replacing an Old Electro-mechanical Relay)

My recommendation: Stick with a simpler configuration that matches the old relay's characteristics. You don't need to enable every digital function.

I once had a client with a 20-year-old GE transformer that needed a protection upgrade. They wanted the latest GE Multilin 845 with all the bells and whistles. I told them: 'Actually, that's overkill and could cause nuisance trips.' We configured it with a simple definite-time overcurrent (50/51) and a ground fault relay (50N/51N). We disabled the negative-sequence overcurrent and the harmonic restraint (for a non-critical load). It worked fine. Trying to apply every advanced algorithm to an old asset is like putting racing tires on a family sedan. (I really should document this rule of thumb.)

Scenario C: Protecting a Transformer in a High Fault-Current Environment (e.g., Near a Large Motor Drive or Generator)

My recommendation: Prioritize the current transformer (CT) selection and saturation characteristics. The relay is only as good as the signal it receives.

In September 2022, I submitted a protection scheme for a 10 MVA unit. The designed GE 850 relay was perfect, but the 1000:5 CTs we selected had an accuracy class that wasn't high enough for the available fault current. Under a through-fault, the CTs saturated, the relay saw a differential current, and it tripped unnecessarily. The result: a $6,000 production loss for the factory in one afternoon. That's when I learned to always calculate the CT saturation voltage (V_s) before selecting the ratio. I assumed the standard values would be fine. They weren't. (Ugh.)

The Decision Tool: Which Scenario Are You In?

If you're reading this and thinking, 'Which one applies to me,' here's a simple decision tree:

• Is this a greenfield project with a high-value transformer? → Follow Scenario A. Invest in the full protection suite.
• Are you replacing an old relay and the transformer is standard? → Follow Scenario B. Keep it simple.
• Is the transformer connected to equipment like large motors or generators? → Follow Scenario C. Check the CT specifications first.
• Is the budget extremely tight? → I'd still recommend investing in the correct GE protection relay as a bare minimum. If a $50 difference per project translates to better protection, that's a no-brainer. Cheap stuff leads to expensive failures.

So, bottom line: The GE transformer itself is a fantastic piece of hardware, but the protection setup is where the engineering lives. Don't assume the defaults are right. (Prices as of January 2025; verify current catalog specs).