Get this new approach to evaluating Voltage Sags

[Sreemant_Roy]

Sharing extracts from a recent Technical Article authored by Jon Bickel, P.E., Power Applications, Schneider Electric.

The paper discusses the impacts of voltage events such as sags / dips and short interruptions related to power quality issues. Whether originating at the utility or inside an end user’s facility, it is important to understand operational impacts, and to differentiate nuisance events from disruptive events. Ascertaining the level of impact from voltage events allows easier prioritizing of alarms, creating and trending historical effects from disruptive perturbations, and determining locations and sizes of mitigation equipment.

Traditionally mid/high-end metering devices have had the ability to capture short duration rms variations such as voltage sags/dips for decades. These devices typically provide a time-series image of voltage perturbations (e.g., sags/dips) that include all metered voltage and current phases.

When analyzing voltage events, there are several questions to be answered including:

What type of voltage event was it?

How long did the voltage event last?

Was the source of the voltage event internal or external 

What caused the voltage event?

Is this voltage event a reoccurring issue?

and most importantly…WAS MY OPERATION IMPACTED?

The figure below illustrates the normal voltage range (shaded area) that is bounded by an upper and lower alarm threshold. The rated or nominal voltage is also shown.

When the measured voltage level goes below the lower alarm threshold (e.g., see the voltage event shown in the figure below) an alarm will be initiated by the metering device.

The event alarm data typically includes information such as worst magnitude of the event, duration of the event, start time and date of the event, alarm type, and (depending on the device) a waveform capture of the three-phase voltages and currents during the event. Each of these event parameters provide important clues to troubleshoot the voltage event and answer the questions listed above.

However, one very important question is not directly addressed: Did the voltage event impact the load? To determine whether the voltage event resulted in a load impact, a more thorough analysis of the waveform capture must be performed.

Let’s look at Load loss

Figure – 1 below illustrate a voltage sag event on a three-phase, 60 Hertz, 347/600-volt wye-configured system. The event shown in this example appears to have been caused by an upstream fault.

Figures 2&3 below are derived from the event waveform data captured in the figure -1 above. The voltage event lasted approximately 11 cycles (or about 180 milliseconds) in duration, and the worst-case voltage deviation during the event was approximately 53% of the nominal voltage (or 184 volts) on Phase C as shown in figure 2 (rms voltage) below.

Figure 3 illustrates the rms real power data throughout the voltage event and is the data with the most relevance for troubleshooting the event’s impact. Before the event began (pre-event), the total real power consumed by the downstream load(s) was approximately 1,458kW.

The voltage event began when t ≈ 0.07 seconds and is assumed to have concluded once the voltage recovered to its normal operating range (t ≈ 0.26 seconds). Upon the conclusion of the event (i.e., post-event), the total real power consumed by the downstream load(s) was approximately 544kW.

Therefore, the pre-event versus post-event total power flow to the load(s) decreased by about 914kW (or 63%) as a result of the voltage sag event.

The conclusion is the voltage event caused 914kW (or 63%) of load to de-energize (drop offline), likely resulting in a significant impact to the facility’s operation.

Conclusion

Globally, voltage events such as sags/dips and short interruptions are the biggest contributor to losses related to power quality issues. Voltage events can be external (e.g., originate on the utility) or internal (e.g., originate inside the end-user’s facility), anticipated (e.g., starting a large load) or unpredictable (e.g., a system fault), impactful (e.g., loads de-energize) or inconsequential (e.g., system continues to operate with no issues).

Recognizing the existence of voltage perturbations and characterizing their properties (e.g, worst magnitude, duration, etc.) is not sufficient; it is important to understand the operational impact to differentiate nuisance events from disruptive events.

Ascertaining the level of "Operational impact" from voltage events (regardless of their origin) facilitates easier prioritizing and filtering metering system alarms, creating and trending historical effects from disruptive perturbations, and determining locations and sizes of mitigation equipment like Schneider Electric’s PowerLogic Dynamic Voltage Restorer ( DVR), is the "Key" to resolving a client's PQ Challenge.

Discover more about PowerLogic DVR HERE

[Charles_Murison]

Thank you for sharing this information.

Regards,

Charles

[Sepehr_Zarrabi · Accepted solution]

Great article, which makes a very good point: regardless of the severity of the event, how did the loads / processes get impacted?

The "load loss" calculation discussed in this article is performed automatically by the PME (Power Monitoring Expert) software (image below).