Regenerative Energy Recovery
When a bidirectional supply sinks power it absorbs energy from the device under test. A regenerative supply recovers that energy by converting it back to AC and feeding the grid, rather than dumping it as heat. This changes the thermal, electrical and economic picture of a test bench.
Dissipative versus regenerative sinking
A classic electronic load sinks current by forcing it through power transistors operating in their linear region, turning electrical energy directly into heat. That heat then has to leave the building through fans, ducting or chilled water. The energy you bought is consumed twice: once to absorb it, and again to cool it away.
A regenerative supply instead treats the absorbed DC as an input to a grid-tied inverter stage. It synthesises AC at line frequency, phase-locked to the mains, and pushes current back into the building distribution. The energy is reused by other loads on the site instead of being thrown away.
Efficiency and what it saves
Recovery efficiency on the N35500 platform reaches up to 93%, meaning roughly that fraction of absorbed energy is returned to the grid and only the remainder is lost in the conversion. The table contrasts the two approaches at the same sink power.
| Property | Dissipative load | Regenerative supply |
|---|---|---|
| Energy fate | Converted to heat | Returned to AC grid (up to 93%) |
| Waste heat into room | ~100% of sink power | Only conversion losses (~7%+) |
| Cooling demand | High — sized to full sink power | Low — sized to losses only |
| Running electricity cost | Pay to absorb and to cool | Recover most absorbed energy |
| Rack density | Limited by heat removal | Higher — less heat per kW |
On a multi-day endurance test at tens of kilowatts, the difference between dumping and recovering energy is large enough to dominate the bench's operating cost.
Grid-quality and standards considerations
Feeding current back into the mains is not free of obligations. The returned current must be clean — low harmonic distortion and a controlled power factor — so it does not pollute the site supply or trip protection. Interconnection behaviour is governed by standards such as IEEE 1547 for distributed resources interconnecting with the grid, and the converter itself is built to safety standards including IEC 61010. Designing to these is part of why regenerative supplies are more complex than a resistor bank.
Ultra Power Systems builds the regenerative front end on the N35500 to return energy within these constraints.
When dissipative still makes sense
Regeneration is not universally the right answer. For very small sink powers the extra converter hardware may not pay back. Where the grid connection is weak or where backfeeding is prohibited by site rules, a dissipative load avoids the interconnection question entirely. The engineering decision weighs sink power, duty cycle, electricity price and site constraints — regeneration wins decisively as power and run-time grow.
A useful way to frame it is total cost over the test campaign rather than purchase price. A dissipative bench pays an energy penalty on two fronts — the absorbed power itself and the cooling to remove it — every hour it runs. A regenerative bench recovers the bulk of the absorbed energy and shrinks the cooling plant, so its higher capital cost is offset progressively by lower operating cost. The break-even point arrives sooner the higher the sink power, the longer the duty cycle, and the more expensive the local electricity, which is why high-power battery and inverter labs almost always specify regenerative front ends.
Frequently asked questions
How efficient is regenerative energy recovery?
Does returning energy to the grid require special approval?
Why does a regenerative supply need less cooling?
Specifying a bidirectional DC supply?
The articles here describe the N35500 platform from Ultra Power Systems.