Transcript

Introduction

Thank you for joining today’s Ask a Pro webinar. My name is Wesley Houck. I’m an electrical engineer here at PMI. Today, we’ll be covering our white paper on modern flicker analysis using PQ Canvas and Merlin.

Voltage flicker is one of the most common power quality complaints reported by utility customers. It refers to visible changes in lighting intensity caused by fluctuations in supply voltage. As modern electrical systems evolve, we’re seeing more loads that create these fluctuations, such as arc furnaces, welders, large motor starts, and power electronics.

To measure flicker objectively, we use a standardized tool called the flicker meter, defined in IEC 610004-15 and adopted by IEEE 1453. This produces key metrics, like instantaneous flicker level, short-term flicker severity, and long-term flicker severity. This paper focuses on flicker fundamentals, how the flicker meter works, and how modern tools like PQ Canvas and Merlin improve analysis.

What Is Flicker?

Flicker is defined as the visible change in brightness caused by fluctuations in supply voltage. It’s important to understand that flicker is based on human perception. The electrical issue is voltage fluctuation, but flicker is how we see it.

The human eye is most sensitive to light modulation between 5 and 15 hertz, with a peak sensitivity around 8.8 hertz. Because of this, even very small voltage changes, sometimes less than .5%, can be noticeable.

Causes of Voltage Fluctuations

Voltage fluctuations are typically caused by a changing load current flowing through system impedance. As current changes, voltage drop changes as well. This creates modulation in the supply voltage.

Typical sources include arc furnaces, welders, large motor starts, heating systems, and power electronics. Even residential loads like HVAC systems can contribute. Because flicker depends on perception, we need standardized measurement methods rather than relying on voltage alone.

The Flicker Meter

To address this, the IEC developed the flicker meter. This is a signal processing system that simulates how a lamp and the human eye respond to voltage fluctuations. It produces three key outputs:

• Instantaneous flicker level, or IFL
• Short-term flicker severity, or PST, which is calculated over 10 minutes
• Long-term flicker severity, or PLT, calculated over two hours

These standardized values allow consistent comparison across different systems and devices.

Flicker Signal Flow Block Diagram

On figure one, it shows the flicker signal flow block diagram. This diagram illustrates how a flicker meter processes the voltage signal step by step.

1. In the first block, the input voltage is normalized to a reference value, ensuring consistent measurement. This stage also produces the relative voltage change, the delta V over V.
2. In block two is the demodulation. This removes the fundamental, or 50 or 60 hertz signal, and isolates the modulation caused by voltage fluctuations.
3. Block three filters simulate how sensitive the human eye is to flicker, especially around 8.8 hertz where severity is the highest.
4. In block four, this module shows how flicker is perceived over time, accounting for persistence of vision.
5. In block five, this stage calculates the standardized metrics that we discussed earlier, the IFL, the PST, and PLT.

Overall, this process converts raw voltage fluctuations into values that represent human perception of flicker.

Analyzing Flicker Data with PQ Canvas

For modern power quality monitors that record flicker data, today this data is analyzed using PQ Canvas, a cloud-based platform. With PQ Canvas, engineers can examine PST strip charts, instantaneous flicker levels, voltage RMS variations, and load current correlations.

A key technique is comparing flicker levels with load current. As shown in figure two on page three, we can compare instantaneous flicker levels with the load current. If flicker spikes line up with the current spikes, the source is likely downstream of the monitor. If the flicker occurs without current changes, the source is likely upstream. In this example shown, the flicker source is likely upstream. This type of analysis helps engineers determine both severity of the flicker and where it is.

PQ Canvas Recording Example

We can also see this here in this example on PQ Canvas. This recording is a Bolt set up in Y configuration. If we go here to the PST flicker strip charts, we can see that anything above one on this strip chart is violating the IEEE 1453 standard.

This one doesn’t really count, because this is caused by a storm, so we can ignore that and zoom in on the rest of the strip chart data. In here, you can see that for the most part it’s below one on the PST level, but in a few points you can see that it gets just above one. So it does violate it, but just slightly.

IEEE 1453 Compliance Limits

The IEEE standard defines recommended limits to minimize customer complaints. Typical values are:

• Short-term flicker severity, PST equal to or less than one
• Long-term flicker severity, PLT less than or equal to .8

These limits are based on human perception studies. About half the people report noticeable flicker when the PST reaches 1. Utilities often evaluate compliance statistically, ensuring these limits are only exceeded a small percentage of the time.

AI-Based Analysis with Merlin

PMI also offers an AI-based tool called Merlin that works on top of PQ Canvas, which I will show you. I’ll run Merlin on the recording that we viewed earlier when we looked at the PST flicker.

Traditionally, analyzing flicker required manual inspection of the voltage and current and flicker data, which can be time-consuming. To improve this, Merlin was developed as an AI-based analysis tool integrated with PQ Canvas. Merlin automatically analyzes power quality recordings and identifies patterns associated with flicker. It evaluates relationships between voltage fluctuation, flicker levels, and load current.

Looking at figure three on page three, we see an example of Merlin’s output, but I will get more into this here soon. In this case, flicker is rated nine out of 10 on the Compliance Severity Scale, indicating it is significantly out of compliance.

Merlin can also identify:

• Correlations between current spikes and flicker
• Periodic load behavior
• Patterns consistent with known flicker sources

This allows engineers to quickly focus on the most important parts of the data instead of manually reviewing everything. It also provides summaries, estimated causes, and helps generate reports. However, it’s important to note that Merlin does not replace engineering judgment, it enhances it.

Merlin in PQ Canvas Demo

If we go back over here to PQ Canvas, get out of the strip chart, and come up here to the Merlin overview, you can see what we saw earlier. It was under one for the most part, but a few times it got just above one, so just a little bit out of compliance. Here it says borderline long-term flicker from load-driven voltage dips.

If we go into here, it’ll give you a full executive summary that you can use for reports to customers. You can send this out so the engineers do not have to spend time coming up with all this data and analyzing the report. It’s all written out here for you.

If you want to get deeper into the analysis, you can come down here or scroll down to Flicker Analysis. Here, it really dives into what Merlin saw when looking through all the data associated with this recording. Here’s an in-depth analysis of all of that.

We can come back to the top dashboard and come over here to the Compliance Report. This focuses strictly on the IEEE 1453 compliance of this recording. You can dive into here, and it indicates mild long-term flicker non-compliance on one of two 120 volt voltage channels. So it’s saying that for the most part, like we said, it was in compliance.

Conclusion

In conclusion, flicker remains a common power quality issue because even small voltage fluctuations can produce visible lighting disturbances. The flicker meter provides a standardized way to measure flicker using IFL, PST, and PLT. Modern tools like PQ Canvas make it easier to analyze and visualize flicker data, and with the addition of AI tools like Merlin, engineers can diagnose flicker problems faster and more efficiently. By combining these technologies, utilities can better understand and respond to flicker in modern electrical systems.

Q&A: Identifying Flicker Source Direction

In this example, I took this out of a different recording that I saw. Let’s zoom in a little bit. Here we have the RMS Current Max on the top, and then we have the IFL on the bottom.

If you see a correlation between IFL spikes and the IFL strip charts, and if there’s corresponding spikes in the current for every single one of these spikes, then that would mean that the cause of the flicker is actually in-house or downstream. Whereas in this example, there’s no correlation between these spikes and the current. So that means that the flicker cause is somewhere upstream from the recording device.

Closing

Thank you for attending today’s webinar. If you have any questions, you can call us at 800-296-4120 or email support@powermonitors.com. Have a great evening.