,
Abstract
The job of any power monitor is to record all interesting data, and leave unrecorded the vast majority of boring, unremarkable data. The tricky part for a monitor is deciding which events are important. This is a problem of data reduction. A recorder that captured every 60 Hz waveform during a week’s recording would never miss an event, but would present the user with billions of useless cycles. Conversely, a recorder whose thresholds are set incorrectly may not record anything. Staying somewhere between these two extremes involves a balance of thresholds, settings, and record types. The monitor will see an enormous amount of data on its voltage and current inputs– the Revolution sees over 432 billion samples per day! Ideally, all this data is reduced to a small report which just shows the important events and measurements. The sifting of data into specific record types accomplishes this task.
Part 1 of this series covered all the non-triggered PMI record types (e.g. stripcharts, daily profiles, histograms, etc.) The next parts in this series cover triggered record types. These range from slowly triggered events such as outages and loose neutral to high speed events such transient capture, triggered from single 1 microsecond excurions. Slower record types triggered by one second data (Outages, Abnormal Voltage, Loose Neutral) are covered in this whitepaper. Cycle-based events (Significant Change, Event Capture, Flicker) will be in the next part in the series, and the final paper will cover sample-based events such as waveform capture, ITIC/CBEMA, and transient capture.
Triggered Record Types
Data record types can be divided into two classes. The first is event driven. These record types are triggered by a combination of triggering logic and adjustable thresholds, often voltage-based. If a trigger condition is never met, nothing is recorded for that record type. As more triggers occur, more records are collected for that record type. The advantage of this class is that nothing is recorded unless something happens. In the ideal case, no problems occured, so nothing was recorded, and no data analysis is necessary. If a trigger did occur then the monitor logged the event for later analysis. This is a powerful data-reduction tool which can reduce huge amounts of data into a few small records containing all the significant events. The disadvantage is that success depends on good thresholds and settings. A threshold that is too tight will cause the recorder to log records that aren’t really worth analyzing. These extraneous records often hide the (hopefully) few important ones. A threshold that is too loose will cause the recorder to ignore important disturbances. Although it is often possible to use regulatory limits or other known standards to set threholds, this can be a chicken-and-the-egg type problem: sometimes you need to know something about the disturbance before you can set proper thresholds to capture it. Despite these potential pitfalls, triggered record types are powerful tools in powerline monitoring. They are most useful for capturing voltage disturbances and power quality problems. Triggered record types include Waveform Capture, Transient Capture, ITIC/CBEMA, Event Change, Significant Change, Flicker, Loose Neutral, Power Outage, and Abnormal Voltage.
Non-triggered Record Types
The second class of record types is not event driven. These record types are always logging data regardless of how interesting or important the data is. The classic example is a paper stripchart which continuously logs data. These record types include Stripcharts, Daily Profiles, Histograms, and Energy Usage. These record types are reviewed in part 1 of this series.
Remote Access and Memory Usage
Traditional PQ recorders have a finite amount of memory for data storage. In the past, managing this limited memory involved a tradeoff between different data types, thresholds, etc. to help insure that the essential data was preserved. Modern recorders such as the Revolution with up to 1GB of on-board memory greatly eased the burden, but still present a finite recording memory. The introduction of Ethernet and cellular remote access to the Revolution eliminates this problem altogether in two ways.
First, the PQ memory is effectively infinite with the use of automatic scheduled downloads in ProVision. For example, the Revolution may be configured to fill its internal memory with a one week recording. If a weekly automatic download via cell modem (or Ethernet) connection is scheduled in ProVision, the Revolution is downloaded and reset each time, effectively offloading the bulk data storage from the Revolution to the downloading PC. With this setup, detailed high resolution PQ recordings may be continued indefinitely. Scheduled downloads are especially useful with triggered record types, where the amount of memory used (and thus recording time) depends on actual line conditions, and can be difficult to predict.
Second, the Revolution can be configured to trigger email or text message alerts based on exceedances such as voltage, current, or power out of limits, or waveform capture triggers. These alerts work through PMI’s Canvass cloud-based system, and provide instant notification of a PQ event. With that notification, ProVision can be used to download the Revolution to retrieve the full data set, without waiting for the next scheduled download, or for field retrieval of the Revolution. These alerts actually remove the need for indefinite recording in some situations – the PQ data can be examined as soon as the event happens, rather than waiting for a long recording to end. This alert-based monitoring actually reduces the amount of field data needed for an investigation and shortens the time required to identify and start resolving a problem.
Power Outage
The Power Outage report lists the date and time of all outages during the recording session. An outage is defined by the recorder to be a voltage sag below 80 volts lasting for at least 1/3 of a second. Only channel one’s voltage is used to trigger an outage. The beginning and end of the outage are timestamped. In the report, the duration is also given along with the total number of outages and the total outage time.
If the recorder has battery ride-through capability, it will continue to record Histograms, Stripcharts, etc. during the outage. If there is no battery, or if the battery runs down, the recorder loses power and stops recording. When power is restored the recorder records the end of that power outage and resumes recording normally.
A power outage often triggers Waveform Capture, which may help reveal the cause of the outage.
Examples
A typical Power Outage report is shown in Figure 1. Five outages were recorded over this 18 day period. The best way to use the Outage report is to first check the Header Report (the default report that launches automatically when a new file is opened). If any Outages were recorded there will be a clickable link (as seen in Figure 2, circled in orange). Click this to launch the Outage report, which in many cases will be the most important PQ event in a recording from a customer perspective.
Figure 1. Power outage report showing 5 outages over an 18 day period
Figure 2. Header report showing clickable link of outages (circled in orange)
Abnormal Voltage
The Abnormal Voltage record type shows if the average line voltage moved past a low or high threshold from the nominal voltage. On some older recorders, the low threshold exceedence is indicated by a green LED on the front panel, and the high threshold exceedence by a red LED. When the trigger occurs, the event is timestamped to the nearest second. There is a separate LED and report for each voltage channel.
Trigger Logic
The triggering logic uses a low and high threshold, a nominal voltage, and a trigger duration. The thresholds are added and subtracted to the nominal voltage to find triggering points. If the voltage crosses a triggering point for longer than the trigger duration, an Abnormal Voltage event occurs.
The recorder is initialized with a list of potential nominal voltages (such as 120, 240, etc.), with low and high voltage thresholds for each. The actual nominal is picked by the recorder during the two minute countdown. The average voltage during the countdown is compared to each of the nominals; the closest one becomes the nominal voltage for the entire recording session. There are five standard nominals in the software setup (120, 208, 240, 277, and 480 volts), and two custom nominals. The custom nominals can be set to any voltage. It is possible to enable and disable the standard and custom nominals. For example, if you wanted to force the recorder to use 230 volts as the nominal, the standard nominals should be disabled, and both custom nominals set to 230. If the standard nominals were not disabled there would be a chance for the recorder to pick 240 volts during the two minute countdown, if the line voltage happened to be running closer to 240 than 230 at that time. The nominal is chosen by the recorder separately for each voltage channel.
There are separate high and low thresholds for each of the seven nominal voltages. The applicable thresholds are used once a nominal is selected by the recorder after the two minute countdown. Voltage channels are handled separately; there is a complete set of nominals and thresholds for each. This is useful for situations such as a hot-leg delta, where one voltage channel is at a different voltage, or in a single phase setup where two channels are connected line-to-neutral, and one channel is line-to-line. The recorder will automatically select the correct nominal and thresholds for the different line voltages on each channel.
The last Abnormal Voltage parameter is a trigger duration, in seconds. This specifies how many seconds in a row the voltage must exceed the threshold before the Abnormal Voltage record is triggered.
At the end of each second during the recording session, the recorder compares the one-second average voltage with the nominal and the low and high thresholds. Each threshold actually creates two trip points, one above the nominal and one below. For example, consider a setup where the nominal is 120 volts, the low threshold is 6, and the high 12. The low trip points become 120±6, or 114 and 126 volts. The high trip points are 120±12, or 108 and 132 volts. If the one-second average voltage rises above 126 or falls below 114 volts for longer than the trigger duration, the low Abnormal Voltage trigger occurs. This event is timestamped, and the green LED is lit (if present). If the voltage goes past either high trigger point (108 or 132 volts) for longer than the trigger duration, the high Abnormal Voltage trigger fires. This is timestamped and logged. It’s possible for the low and high triggers to fire at the same time.
The use of one-second average voltages eliminates false triggering due to momentary sags and swells. Abnormal Voltage is designed to trigger for average line voltage exceptions, not sub-second events.
What’s Recorded
When Abnormal Voltage is triggered, the date and time, along with the channel and triggering voltage are recorded. There is a separate listing for each voltage channel, as well as low and high thresholds. Only the first trigger for each threshold is recorded. The nominals selected during the two minute countdown are also recorded.
Typical Settings and Suggested Uses
The Abnormal Voltage report is used to determine whether the voltage drifted outside the thresholds during the recording session. Just as useful is the set of nominal voltages selected from the Abnormal Voltage settings during the countdown period. These nominals are used by several other data types to set baseline voltage values during the recording, and in many cases this function is more important than the Abnormal Voltage report itself. The five built-in nominals cover most circuits, and the two custom slots can be used for non-standard applications.
Usually the Abnormal Voltage report data is used to get a quick read of whether there was any line voltage drift- if so, then other record types such as the Stripchart and Significant Change are used for more information. There will be a clickable link in the Header Report if an Abnormal Voltage event is recorded.
The default threshold settings are at 5% and 10% of the nominal voltage (for example, 6 and 12 volts for the 120 volt nominal). The high threshold must be larger than the low threshold. The two custom nominals are preset at 106 and 230 volts, but should be changed if a different nominal is in use. The default trigger duration is five seconds, and can be set as small as one second, or as large as 255 seconds.
Examples
A sample Abnormal Voltage report is shown in Figure 3. This recording is from a single-phase residential service, resulting in a 120V nominal for each channel. The 5% threshold were exceeded on each leg at separate times, as seen in the report. The RMS Voltage stripchart is graphed in Figure 4. Dashed red lines are shown at the 5% low mark (114V). The dark green trace is the one-cycle minimium, and the light green is the 1 minute average. The minimum dips below the limit several times, but the average does not (although it comes close). The Abnormal Voltage triggering logic, based on one second averages and an adjustable trigger time, nicely caught the voltage change that fell between the single cycle and one minute time periods in the stripchart.
Figure 3. Sample abnormal voltage report from residential single-phase service
Figure 4. RMS voltage stripchart showing 5% low mark (dashed red), one cycle minimum (dark green), and 1 minute average (light green)
Loose Neutral
The Loose Neutral report shows whether the typical symptoms of a loose neutral have occurred. This report is intended for single phase services, with voltage channels one and two connected from line to neutral. Only a two-channel recorder (or a recorder initialized to use two channels) can record a Loose Neutral. The symptom of a loose neutral condition is for one voltage leg to rise in voltage, and the other to fall, with the sum of the two voltages remaining close to twice the nominal voltage. For example if the voltages start at 119 and 121 volts, then move to 105 and 135 volts, a loose neutral is a likely cause: one leg went up, one went down, and the sum is close to twice the nominal (240 volts).
This happens when the load is not balanced and the neutral is disconnected (or has a high impedance). If this condition persists for long enough the Loose Neutral report is triggered.
Trigger Logic
The Loose Neutral logic uses three parameters: duration, range, and difference. These parameters are used to judge whether one voltage leg has risen, and one fallen, while the sum remained the same. The difference is a voltage that specifies the minimum difference between the two legs. For example, if the difference is 16 volts, then there must be at least a 16 volt separation between the two legs. The range is a voltage that specifies how close the sum of the two voltages must be to twice the nominal. For example, a range of 12 volts means that the sum of the two legs must be within 12 volts of twice the nominal voltage. Both the range and the difference conditions must be met for at least the number of seconds specified by the duration. If the duration is set to 5 seconds, then the difference and range conditions must be met for 5 consecutive seconds before a loose neutral is declared. One-second average voltages are used. The nominal voltage is the nominal determined during the two minute countdown by the Abnormal Voltage record type, and is typically 120 volts in a single-phase hookup.
As an example assume the difference parameter is 16 volts, and the range 12 volts, with a duration of 5 seconds. The two line voltages are 119 and 121 volts. Then one leg moves to 128 volts, and the other to 110 volts. The difference between the two legs is 18 volts, which meets the difference threshold. The sum of the two voltages is 238 volts, which is within the required 12 volts (specified by the range value) of twice the nominal (240 volts). If these voltages persist for 5 seconds in a row then a Loose Neutral record will be triggered.
If one voltage leg changes due to heavy loading the range parameter keeps the loose neutral from false triggering. For example, if the voltages start at 119 and 121 volts, then a heavy load to channel 1 causes it to drop to 105 volts, with the other leg still at 121, the difference condition is met (121–105>12), but the range condition is not met: 105+121=226, and 226 volts is not within 12 volts of the 240 volt nominal.
What’s Recorded
The date and time of the loose neutral triggering is recorded along with the voltage on the two channels. Only the first occurrence of a Loose Neutral is recorded; if the conditions are met again, nothing further happens. The Loose Neutral report shows whether the neutral may have a bad connection, not the exact times the connection was made and broken.
Typical Settings and Suggested Uses
The Loose Neutral Report can show the symptoms that may indicate actual loose neutral connection. It is worth investigating if the report is triggered. However, it is possible for the Loose Neutral logic to be fooled. If both legs are equally loaded, then the two voltages will remain the same even if the neutral is removed. This will prevent the Loose Neutral trigger from firing. It is also possible for one leg to rise and one to fall due to grossly different loading, and not from an actual loose connection. Thus it is possible for a Loose Neutral to trigger falsely, when there is no loose connection. Conversely, the voltage movements that trigger Loose Neutral are dependent on unequal loading on the two legs, and if that’s not present a loose neutral could exist without exhibiting any triggerable symptoms. Also, with an intermittent condition a loose neutral could cause voltage events that are too fast for Loose Neutral to catch.
Examples
A Loose Neutral report is shown in Figure 5. The trigger conditions were met at 10:09:59 AM, indicating a possible loose neutral in this residential location. Channel 1 fell to 116V, while channel 2 rose to 135V. The RMS voltage stripchart for this period is shown in Figure 6. Here the min and max traces are both moving wildly, with the averages moving as per the Loose Neutral Report.
Figure 5. Possible loose neutral at 10:09 AM
Figure 6. RMS voltage stripchart with large variations in min (green) and max (red) traces
Conclusion
A quick summary of the second-level triggered record types used by PMI recorders has been presented, describing what’s recorded, any parameters to adjust, triggering logic, and some suggested uses and examples. Abnormal Voltage, Loose Neutral, and Power Outage reports are basic but important tools for PQ surveys and identifying voltage delivery problems. These reports also contain important customer-facing information. Together, these data types are key elements of a full PQ recording.
Chris Mullins
VP of Engineering and Operations
cmullins@powermonitors.com
http://www.powermonitors.com
(800) 296-4120
