One case measured on a Reinigen installation, and four installations reported by the developer of MPTS. Each shows how it was measured and whose installation it was.
A power recorder on the main switchboard, before the unit went in (12–17 June 2024, about five days) and after (21–28 April 2026, seven days). The charts are traced from the recorder’s own printouts.
| Measure | Before | After |
|---|---|---|
| Average power factor | 0.870 | 1.000 |
| Lowest power factor | 0.788 | 0.998* |
| Average reactive power | 101 kVAR | 3.0 kVAR |
| Apparent power above real load | 16% | 0% |
* The recorder’s first April reading, logged as it started, was 0.950 PF and 54.8 kVAR. Every later reading stayed between 0.998 and 1.000 power factor and under about 10 kVAR.
Apparent power above real load: average kVA over average kW. Before, 203 kVA carried 175 kW of real load; after, 116.1 kVA carried 116.1 kW.
Put your own switchboard’s numbers in the Power Factor CheckThe before and after weeks lined up by weekday and clock time. Press Play, or drag across the chart, and read both dials.
0.836
power factor
120 kVA carried for every 100 kW of work
Below 0.90Below 0.90 for 24.5 of 27 hours so far
0.999
power factor
100 kVA carried for every 100 kW of work
0.90 or betterBelow 0.90 for 0 of 27 hours so far
Drag across the chart to scrub. Half-hour averages traced from the power recorder’s own charts; the two weeks line up by weekday and clock time. They ran different loads, June and April, so the replay compares power factor only, never kW or dollars. The April recording’s first reading (0.950, logged as the recorder started on the Tuesday) falls before the window shown.
Thursday 2:30 pm. Before 0.836, below 0.90. After 0.999.
Half-hour averages as a line; the shaded band spans the lowest and highest readings in each half hour.
| Week | Average | Minimum | Maximum |
|---|---|---|---|
| Before (12–17 June 2024) | 0.870 | 0.788 | 0.986 |
| After (21–28 April 2026) | 1.000 | 0.950 | 1.000 |
The April recording’s first reading, logged as the recorder started, was 0.950 PF and 54.8 kVAR. Every later reading stayed between 0.998 and 1.000 PF and under about 10 kVAR.
Before, reactive power stepped between roughly 40 and 180 kVAR with the working day. After, it stayed within a few kVAR of zero around the clock.
| Week | Average | Minimum | Maximum |
|---|---|---|---|
| Before (12–17 June 2024) | 101 kVAR | 23 kVAR | 202 kVAR |
| After (21–28 April 2026) | 3.0 kVAR | −3.6 kVAR | 54.8 kVAR |
The April recording’s first reading, logged as the recorder started, was 0.950 PF and 54.8 kVAR. Every later reading stayed between 0.998 and 1.000 PF and under about 10 kVAR.
Read this before comparing kW
Power-factor correction cannot cut real power by a third on its own. The June week carried summer cooling load, and its overnight load alone ran nearly twice April’s. That is why we don’t count the kW or kVA difference as savings.
| Week | Average | Minimum | Maximum |
|---|---|---|---|
| Before (12–17 June 2024) | 175 kW | 62 kW | 304 kW |
| After (21–28 April 2026) | 116.1 kW | 47.4 kW | 197.4 kW |
These installations belong to Trans Power Company, the developer of MPTS, and come from its case-study material. We did not measure them, and results at another site will differ.
Utility bills and 15-minute interval data, 2016–2019, plus panel instrumentation.
Three MPTS units on a 3,000 kVA transformer; two-week electrical fingerprint.
Instrumentation and utility billing data, continuous production.
An early installation on a 1,200 kVA transformer; monitoring data from the facility’s engineers.
