Precision power quality analysis with 200ms data averaging

11-second data averaging is a fundamental process in power quality (PQ) monitoring and analysis. It’s used to condense vast amounts of high-speed sampled data into manageable, meaningful values. This approach balances the need for detailed event capture with efficient long-term data storage. However, when an event such as a voltage sag occurs, 200ms sampling and averaging offer precise granular data insights to get to the root of the problem faster.

Here’s how data averaging works:

1. Continuous Root Mean Square (RMS) calculation: High-frequency samples are fed into an analog-to-RMS conversion process.
2. User-defined intervals: The user selects the averaging period, which can range from 1 second up to 30 minutes. A common misconception is that the signal is only sampled once per second; in reality, each 1-second RMS value is a composite measurement derived from thousands of individual high-speed samples during that second.
3. Data reduction: The calculated RMS values (e.g., one value per second) are then stored as "trend data". This drastically reduces the data storage requirements while providing a valuable record of voltage and current trends over time.

Data averaging strategies

Power quality analyzers sample the incoming AC signal at very high speeds, often thousands of times per second, significantly exceeding the typical rate for the nominal 50 or 60 Hz fundamental frequency. This high-resolution sampling is crucial for accurately capturing the detailed waveforms of power system events.

However, managing such a massive volume of data over time is a big challenge. Storing raw, high-speed samples for long periods is inefficient and impractical due to storage limitations; you’re bound to run out of space. To reduce the data storage demand, boost memory efficiency, and precisely analyze data, power quality analyzers employ two strategies: constant 1-second averaging with efficient “trend data” conversion, and short periods of 200ms averaging for granular insights into power quality events.

Next, we’ll explain the differences between both methods of data averaging.

Always-on background sampling and RMS averages

Even while operations are seemingly going well, your power quality analyzer is quietly monitoring in the background with 1-second sampling, checking for any irregularities. In the absence of a trigger (based on your chosen predetermined threshold), the instrument continuously processes the high-speed samples to calculate a Root Mean Square (RMS) value over a predefined time interval (also known as data averaging). As it steadily collects samples, the analyzer will produce a time stamp with average, maximum, and minimum values. It is then stored as “trend data” and turned into a graph of valuable voltage and current trends over time. The original raw data is no longer needed, drastically freeing up storage space – a vitally important feature due to the high volume of sampling data and memory limitations.

1-second averaging is more than sufficient for day-to-day operations; it’s unnecessary to constantly monitor power quality at the highest speed possible, and finite memory capacity makes it impractical in practice. When you need a deeper look into the data, you can focus in at higher speeds for a desired amount of time. That’s where the 200ms data averaging feature comes in.

200ms data averaging: Focused power quality insights

While 1-second RMS values are suitable for general applications, faster averaging rates provide greater insight into rapidly changing loads. 

When you need granular data to closely assess a power quality event, the 200ms data averaging feature can provide additional insight for a chosen period. A faster trend rate offers enhanced resolution; 200ms averages produce 5 RMS values per second. This enables you to more accurately capture and characterize rapid responses from systems and the impact of quickly switching loads.

Raw, high-resolution samples are typically only stored for a short duration around specific preset thresholds called trigger events. These triggers indicate the presence of a significant power quality disturbance, such as:

• Voltage sags/dips: Voltage dropping below a set percentage of the nominal value (for example, 90%)
• Voltage swells: Voltage rising above a set percentage (for example, 110%)
• Frequency anomalies: Deviations from the standard frequency

Due to the sheer density of data, constant power quality monitoring at 200ms is neither practical nor necessary. If you do so, you’ll quickly find that you’re running out of memory, and that’s a big problem. 0.2-second averages are a powerful tool, but they aren’t intended for 24/7 monitoring. A common use case includes analyzing sudden changes, such as load steps on generators or large switching loads. It shouldn’t be your default; it’s reserved for times you need to dive deeper into your power quality data.

When using 200ms data averaging, you should collect data anywhere from 1 second up to 30 minutes. Again, due to the high amount of memory required, we don’t recommend using it for over 30 minutes. But when you’re trying to get to the bottom of a power quality event, 200ms averaging is critically useful.

In summary

Data averaging is a vital technique in power quality monitoring that converts massive streams of high-resolution sampled data into practical, trended RMS values. With a power quality analyzer capable of both 1-second and 200ms averaging speeds, efficient long-term analysis and detailed event investigation are possible.