Capacitor Shelf Life and Reforming: Bringing a Stored Driver Back Into Service
An aluminium electrolytic capacitor does not stay new on a shelf. Its electrolyte slowly changes, the oxide layer on the anode can thin where it has not been stressed, and the leakage current rises with storage time. The process is slow but real, and it is the reason manufacturers specify a storage life and a procedure for bringing a long-stored unit back into service. Ignoring it can turn a perfectly good spare into a failed driver on the first power-up.
The mechanism is electrochemistry. The oxide layer that provides the dielectric is maintained by the applied voltage during normal operation. In storage, with no voltage across the capacitor, the layer can degrade slightly, and the leakage current increases. When voltage is applied again the layer re-forms, but the initial leakage can be high enough to heat the capacitor and, in a severe case, to cause venting. Reforming restores the layer gradually by applying voltage through a current-limiting resistor.
Storage life is stated in the datasheet, often two years at a specified temperature, with a note that the clock runs faster at higher temperature. Date codes on the capacitor and on the driver tell you how long the unit has been in the warehouse. Humidity matters too: moisture ingress into the capacitor and into the board affects insulation and solderability, and packaging designed for long storage includes desiccant and a moisture barrier bag.
The practical rule for a warehouse is first in, first out. Spares should be issued in date order, and stock levels should reflect the real consumption rate rather than a comfortable buffer that sits for years. A driver that has been on the shelf longer than its storage life should be reformed or replaced, and the decision should be recorded, because a failed unit at a customer site costs far more than the stock item.
Reforming is not difficult but it is not optional. Apply the rated voltage through a resistor selected to limit the initial current, hold it for a period proportional to the storage time, and check the leakage current before installing the unit. The exact procedure comes from the capacitor manufacturer rather than from folklore, and it should be done with the driver unloaded. Skipping the step and applying full mains directly is the risk the procedure exists to avoid.
For the PV100, the storage conditions and the shelf life are part of the product documentation, and a distributor or a contractor holding stock should follow them. In a project, a driver taken from long storage should be treated as a unit requiring a check before installation rather than as a new part. The cost of a bench check is small, and it prevents the failure that would otherwise be blamed on the product.
The same logic applies to equipment installed but left de-energized for months, such as a seasonal facility or a standby lighting system. A scheduled power-up, even briefly, keeps the capacitors formed and avoids a surprise failure at the start of the season, when the building is occupied and every fixture matters.
Distributors should rotate stock and record storage conditions, because a unit stored in an uncontrolled warehouse in a humid climate ages faster than the datasheet assumes. The date code plus the storage history is the evidence behind a warranty decision, and it is also the information a project needs when a delivery arrives from an old batch.
Load Transient Response: How a Constant-Voltage Driver Handles Sudden Changes
Salt Mist Testing and Coastal Corrosion: IEC 60068-2-52 in Practice
Related Article