Why Your Electrolytic Capacitors Are Leaking Current (And How to Fix It)

Let me tell you about the first time I watched a brand-new capacitor bank turn into a smoking disaster. I was a junior engineer, maybe two years in, and I’d just finished a power supply redesign for a ruggedized industrial controller. Everything tested beautifully on the bench. The ripple was clean. The output voltage was stable. I felt like a genius.

Then we baked it. Not literally—though that came later—but we put it through a standard 85°C burn-in. Within four hours, the leakage current had tripled. By hour eight, the capacitor was drawing over 10 milliamps. It was a warm little brick of failure. My boss, a grizzled veteran with coffee stains on his schematics, just looked at me and said, “You didn't account for the chemistry, did you?”

He was right. I'd ignored one of the most insidious failure modes in all of electronics: high leakage current in electrolytic capacitors. Look—if you've ever seen a power supply hum, a DC rail droop under no load, or a capacitor bulge like it's about to birth an alien, you've seen the consequences. And honestly? Most engineers treat leakage current like a black box. They know it's there, they know it's bad, but they don't understand why it happens.

So let's tear the lid off this thing. Over the next few thousand words, I'm going to walk you through the real, hands-on causes of high leakage current in electrolytic capacitors—the stuff you learn from field failures, not datasheet rumors. We'll cover chemistry, physics, manufacturing defects, and the one mistake that even senior engineers keep making. Seriously, I still see it.

This isn't theory. This is what you need to know to stop your boards from dying.

---

The Silent Killer: Age and Electrolyte Drying

You ever open a 20-year-old electrolytic cap and find that the inside looks like a dried-out creek bed? That's not just a cosmetic issue. It's the primary driver of leakage current in aluminum electrolytic capacitors as they age. And it happens faster than most people think.

The electrolyte inside these caps isn't just there for show. It's the stuff that actually forms the conductive path between the anode foil and the cathode foil. That electrolyte is typically a solvent-based gel or liquid—something like ethylene glycol mixed with boric acid or ammonium salts. Over time, the solvent evaporates. It seeps out through the rubber seal, diffuses through the aluminum can, or simply breaks down chemically.

Here's the kicker: as the electrolyte dries out, its conductivity changes. And not in a good way. The local electric field inside the capacitor becomes uneven. You get hotspots where the field concentrates. Those hotspots stress the dielectric oxide layer (that thin aluminum oxide film that does the actual insulating). When the oxide layer gets stressed, it breaks down in tiny spots, and suddenly you've got a path for DC leakage current that wasn't there before.

The Chemical Decomposition Cycle

Let's get a little nerdy for a second, because this is where the real magic happens. The aluminum oxide layer (Al₂O₃) on the anode is normally a fantastic insulator. It's formed during a process called anodization, which applies a voltage to grow this crystalline oxide. But that oxide isn't static. It's in constant chemical equilibrium with the electrolyte.

When the electrolyte dries out or changes composition, that equilibrium shifts. The oxide starts to dissolve in some spots and grow thicker in others. You get a patchy, uneven dielectric. And wherever the oxide is thin, you get excessive leakage current.

I've seen this in capacitors that were only five years old but stored at 60°C ambient. The electrolyte was practically gone. The leakage current had climbed from 3 microamps to over 500 microamps. That's a catastrophic increase by any standard.

Thermal Runaway: When Leakage Becomes a Feedback Loop

Here's where it gets scary. When leakage current in an electrolytic capacitor goes up, it generates heat. That heat accelerates the drying of the electrolyte. Which makes the leakage current go up further. Which generates more heat.

You see where this is going.

This is called thermal runaway, and it's why you sometimes find a capacitor that's literally split open after a few hours of operation. It wasn't a voltage spike that killed it. It was a slow, creeping increase in leakage current that started months or years ago. The thermal runaway turned that slow creep into a sudden explosion.

Look—if your capacitor is running more than 10°C above ambient, and the leakage current is climbing, you have a time bomb. Don't design around it. Replace the part or redesign the thermal path.

---

Electrical Abuse: Voltage, Ripple, and The Sneaky Culprit

Alright, let's talk about the second major cause of high leakage current in electrolytic capacitors: electrical stress. This is the stuff that happens when you push a capacitor beyond its comfort zone. And believe me, most engineers do it without realizing.

I once consulted for a company making LED drivers. They were using 50V-rated capacitors in a circuit that theoretically saw 48V. That's within spec, right? Wrong. The circuit had a small inductive kick during startup that briefly hit 54V. Not every time, but often enough. After a few thousand cycles, the capacitors started showing high leakage. A few more cycles, and they failed outright.

The oxide layer in an electrolytic capacitor has a maximum voltage it can withstand. Exceed that, even momentarily, and you punch microscopic holes in the dielectric. Those holes don't heal. They grow over time as current flows through them. The result is a permanent increase in leakage current.

Overvoltage Stress and Dielectric Breakdown

When you apply a voltage higher than the capacitor's rated value, you're essentially forcing current through the oxide layer faster than it can support. The oxide breaks down at weak points. Those weak points become localized hotspots. The heat from those hotspots can actually recrystallize the oxide, making it more conductive in that spot.

It's not a short. It's worse. It's a semi-conductive path that draws current continuously. And once that path forms, it's permanent. No reforming process can fix it. I've tried. It doesn't work.

Here's a quick checklist of conditions that cause this:

- Short-duration voltage spikes from inductive loads (motors, relays, switching converters) - Marginally-rated capacitors where the operating voltage is >90% of the rated voltage - Transient overvoltages during power-up inrush events - Reverse voltage (even a few hundred millivolts can damage the oxide if sustained)

Seriously, that last one is a killer. Electrolytic capacitors are polarized for a reason. The oxide layer only forms on the anode. If you reverse the voltage, you start forming oxide on the cathode foil. That destroys the original oxide structure and causes massive, irreversible leakage current flow.

Ripple Current and Internal Heating

Let me ask you a question. When you look at a capacitor's datasheet, do you actually check the maximum ripple current rating? Or do you just pick a capacitance value and hope for the best?

I'd say 60% of the engineers I've worked with ignore ripple current. And it's one of the most common causes of high leakage current in electrolytic capacitors.

Ripple current generates heat due to the capacitor's equivalent series resistance (ESR). That heat raises the internal temperature. And as we already discussed, higher temperature accelerates electrolyte drying and oxide degradation. The ripple current itself isn't directly causing leakage—it's the heat from the ripple that does the damage.

If your capacitor is running at 85°C internal temperature (not ambient), its life expectancy drops by half for every 10°C rise. That's not a rough estimate. That's the Arrhenius law in action. The leakage current will increase slowly at first, then dramatically as the electrolyte degrades.

---

Manufacturing Defects: The Ones That Slip Through

You'd think after 100 years of capacitor manufacturing, they'd have it figured out. And most of the time, they do. But I've seen enough bad batches to know that manufacturing defects are a real, ongoing problem.

I recall a specific incident from about five years ago. A major capacitor manufacturer had a contamination issue in their electrolyte production line. A batch of electrolyte had an unusually high concentration of chloride ions. Those chloride ions catalyzed the breakdown of the aluminum oxide layer. The result? Capacitors that showed normal performance for the first week, then started exhibiting high leakage current that increased exponentially over time.

We spent three weeks troubleshooting a power supply that would fail after exactly 200 hours of operation. Every time. It was maddening. Finally, we realized the capacitors were the common denominator. We replaced them with a different batch from the same manufacturer, and the problem vanished.

Contaminated Electrolyte

The most common manufacturing defect involves electrolyte contamination. The electrolyte is a chemical mixture, and even tiny amounts of impurities can cause problems. Chlorides, sulfates, and heavy metal ions are the usual suspects. They attack the oxide layer directly, creating conductive paths that never heal.

There's no easy way to detect this during incoming inspection. Standard leakage current tests at room temperature might pass. The contamination only becomes apparent under temperature and voltage stress. If you're designing for high-reliability applications, consider doing a burn-in test on a sample batch before committing to production.

Poor Seal Integrity

This one is subtle. The rubber seal at the base of an electrolytic capacitor is supposed to keep the electrolyte in and the atmosphere out. But if that seal is defective—if it has a hairline crack, a molding void, or simply isn't compressed enough during assembly—the electrolyte starts evaporating from day one.

I've cut open capacitors that were only six months old and found them half-empty. The seal looked fine from the outside. But under a microscope? There it was: a tiny gap where the rubber didn't quite meet the aluminum case.

You can't test for this non-destructively. But you can look for patterns. If multiple capacitors from the same reel show excessive leakage current after a few months in the field, suspect a seal issue. The manufacturer needs to know.

---

Environmental Factors: The Hidden Accelerators

Even if your capacitor is perfectly manufactured and properly operated, the environment can still kill it. And I'm not talking about obvious stuff like rain or dust. I'm talking about conditions that creep up on you.

Humidity and Moisture Ingress

Aluminum electrolytic capacitors are not hermetic. They breathe. The rubber seal allows small amounts of moisture to enter and exit over time. In high-humidity environments (think tropical climates, humidifiers, outdoor gear), that moisture can condense inside the capacitor.

Water in the electrolyte alters its conductivity and pH. A change in pH can dissolve the oxide layer or cause it to grow in uneven patches. The result? You guessed it—high leakage current.

I worked on a project for marine electronics once. The equipment was supposed to be sealed, but condensation formed internally anyway. Within two years, every electrolytic capacitor in the unit had tripled its leakage current. We switched to film capacitors for the critical paths. That fixed it.

Low-Temperature Effects

This one surprises people. Cold actually increases leakage current in some cases. Here's why: at low temperatures, the electrolyte becomes more viscous. Its ionic mobility drops. The oxide layer's repair mechanisms slow down. If you apply voltage at -20°C, any existing defects in the oxide don't self-heal as effectively. Those defects grow, and leakage current increases permanently.

When the capacitor warms back up, the damage is already done. The leakage current stays high.

Mechanical Stress and Vibration

This is a niche one, but it comes up in automotive and aerospace applications. Physical vibration can cause the internal foil rolls to shift slightly. If the foils rub against each other or against the case, they can scratch the oxide layer. Those scratches become leakage current paths.

The fix is simple: use capacitors with robust internal construction, or add mechanical support (like conformal coating or mounting clips) in high-vibration environments. Don't assume the capacitor can handle the vibration just because it's rated for it. Test it in your actual mechanical setup.

---