A lamp can dim, a contactor can drop out and a screen can restart during very different supply events. The equipment symptom alone does not say whether voltage was reduced or effectively interrupted.
This guide follows two fictional RMS-voltage traces through retained voltage, duration, phase aggregation, equipment response and evidence for cause.
The numbers teach classification and transfer. They are not a power-quality survey, contractual finding or safe-work instruction.
The practical habit is to preserve a small evidence receipt beside every answer. It names the object, reference condition, units, observation method, transformation and the nearest tempting claim that still does not follow.
Each changed-context check alters one consequential condition while keeping the remaining basis fixed. This makes transfer visible: the reader can predict which field should change, which should remain stable and what new evidence would be required.
The article uses five routes so readers can enter by task, but the mechanism is cumulative. Definitions establish the objects, the worked case carries the arithmetic, evidence checks test rivals, transfer cases change the context and the final route repairs common shortcuts.
A useful comparison begins with a ledger rather than a conclusion. Put the two candidate quantities in separate columns, write the numerator, denominator, reference condition and observation method for each, and leave an explicit blank when a field is unknown. Then carry one fictional record from raw observations to the narrow result. This prevents a familiar percentage, label or instrument display from borrowing certainty from a different calculation. It also makes later correction cheap: one field can change without rewriting the evidence history or silently moving the boundary of the claim.
The examples deliberately include a near-boundary case, a changed denominator and a changed physical or operational condition. These are not decorative variations. They reveal whether the reader has learned the mechanism or only memorised the first arithmetic. Before accepting a conclusion, predict the direction of change, calculate it, and explain any disagreement between prediction and result. When the answer depends on a standard, contract, laboratory method or browser rule, preserve the exact version and keep the general teaching principle separate from the local decision.
For practical use, keep three layers visible: what was directly observed, what was calculated from those observations, and what action or interpretation is being proposed. Evidence can support the first two layers without automatically justifying the third. A good report therefore states the nearest rival explanation, one check that would distinguish it, and one limit that remains after the check. This structure keeps the writing warm and readable while giving another person enough information to repeat the reasoning, challenge an assumption and transfer the method to a new case.
Before leaving the worked route, perform a reconciliation check. Confirm that units cancel as expected, totals and time intervals close, every derived value points back to a stored input, and every linked article answers a materially different question. Then read the conclusion without the surrounding explanation. If it could be mistaken for a universal rule, add the missing condition directly to the sentence. This last pass catches many errors that arithmetic alone cannot reveal: a shifted reference, a hidden exclusion, an unmatched specimen, an unvalidated conversion or a metric used outside the range in which it was defined.
This specialist guide belongs to the Utilities root. It also connects to The Gold Standard of Education | Learning That Transfers into Life and the How X Works hub. The worked cases are original teaching examples; their numbers are stipulated so that the reasoning can be checked.
eduKateSG · Utilities
Find your next learning step
Read the short answer first: calculate retained RMS voltage against the declared reference, pair it with duration, then apply the stated threshold rule. Keep equipment effects and suspected cause in separate fields.
ROUTE 1 · CHAPTERS 1–4
Define the quantities
Name the quantities, reference conditions and reporting rules.
ROUTE 3 · CHAPTERS 9–12
Audit the evidence
Check instruments, algorithms, assumptions and rival explanations.
ROUTE 5 · CHAPTERS 17–20
Report and continue
Repair shortcuts, answer questions and follow verified routes.
Chapter contents
Define the quantities · Chapters 1–4
1. Name the two voltage events
2. Use per-unit voltage before naming the event
Work the mechanism · Chapters 5–8
5. Work the central waveform ledger
Audit the evidence · Chapters 9–12
10. Audit trigger and hysteresis
Transfer the reasoning · Chapters 13–16
13. Transfer to a motor-start event
14. Transfer to automatic reclosure
Report and continue · Chapters 17–20
18. Write the supported event conclusion
A voltage sag is a temporary reduction in RMS voltage while the supply remains present. An interruption is an event in which the measured RMS voltage falls below the interruption threshold. Both magnitude and duration belong in the event label.
Write the evidence in the order a reader can audit: the Utilities quantity, its reference object, the observation or browser rule, the transformation, and the decision being made. Keeping those fields visible prevents a familiar label from hiding a changed basis.
On a 230 V nominal supply, a monitor records 161 V RMS for 0.20 s. The retained voltage is 0.70 per unit, so the trace is a reduced-voltage event rather than a complete loss in this teaching case.
First-principles test: ask what physical, observational or algorithmic operation produces each value in this chapter. Then ask whether the second value could change while the first stays fixed. If it can, the two labels are not substitutes. For Utilities, this counterfactual is often clearer than a definition learned in isolation.
Keep the 0.20 s duration but change the retained voltage from 0.70 to 0.05 per unit. The classification crosses the stipulated threshold even though the duration does not change.
Evidence check: another reader should be able to reproduce the narrow result from the stored definitions, inputs and intermediate steps, then identify the nearest stronger claim that the evidence does not support.
Thresholds, aggregation rules and duration classes depend on the declared standard and instrument setup.
CHAPTER 2 OF 20 · Define the quantities
2. Use per-unit voltage before naming the event
Per-unit voltage divides measured RMS voltage by the declared reference voltage. This makes magnitude comparable across nominal voltages while preserving the reference in the record.
Trace the mechanism before interpreting the result. For Utilities, mark which value belongs to the source, which belongs to the measurement or layout algorithm, and which is a reported summary. A calculation is useful only when those roles remain attached.
Compute 161/230 = 0.70 per unit. Record 230 V as the reference and do not report 70% without saying whether it is retained voltage or a reduction.
Practical transfer: build two records that share the headline number but differ in the condition discussed here. If the correct interpretation changes, the omitted condition is decision-relevant and belongs in the report. This test turns a vocabulary distinction into a reusable Utilities checking habit.
Move the same 161 V reading to a 400 V reference. It becomes 0.4025 per unit, so the interpretation changes because the reference changed.
Practical check: keep one counterexample beside the worked case. If the wording fails when that one condition changes, narrow the sentence until it describes only the evidence actually in hand.
Nominal, declared and pre-event voltages are not automatically interchangeable references.
CHAPTER 3 OF 20 · Define the quantities
3. Separate retained voltage from voltage reduction
A retained voltage of 0.70 per unit means the waveform retains 70% of the reference RMS value. The reduction is 0.30 per unit. Confusing the two reverses the severity statement.
Use a compact decision table with columns for condition, input, operation, result and limit. This makes the comparison reproducible and exposes where a missing field would force an assumption instead of an observation.
For the 161 V case, write both fields: retained voltage 70%; reduction 30%.
Misconception check: a plausible output does not prove the right reference, denominator, ancestor, spectrum or observation rule was used. Reconstruct one intermediate step and one changed-context prediction. If either cannot be stated, keep the result provisional rather than strengthening its wording.
Change retained voltage to 0.40 per unit. The reduction becomes 0.60 per unit without a second measurement.
Transfer check: repeat the reasoning with one consequential condition changed while holding the rest fixed. State what changes, what stays comparable and why the original label alone was insufficient.
Percent language must name its numerator and reference.
Magnitude alone cannot distinguish a brief sag from a sustained undervoltage, and duration alone cannot distinguish a sag from an interruption. The event receipt therefore keeps onset, recovery, minimum RMS and duration together.
Preserve the raw inputs beside the derived result. If a later reader changes one condition, the original Utilities reasoning should still be recoverable without reverse-engineering a rounded headline or an unlabeled screenshot.
The central trace begins at 10:15:20.000, falls to 0.70 per unit and recovers at 10:15:20.200, giving 0.20 s.
Evidence design: retain both a positive case and a near-miss. The positive case shows when the rule applies; the near-miss changes one consequential condition and shows when the conclusion fails. Together they teach the boundary of the Utilities mechanism better than another unlabelled example.
Extend recovery to 80 s while holding magnitude fixed. The same voltage depth may fall outside a short-duration sag category.
Audit check: retain the source version, units, settings, exclusions and rounding rule. A future correction should be able to alter one field without rebuilding the entire explanation from memory.
Boundary timing depends on the windowing and threshold-crossing method.
Use one-cycle RMS windows for the fictional ledger: pre-event 1.00 per unit, event minimum 0.70, 10 cycles below 0.90 at 50 Hz and recovery to 0.98. Ten cycles at 50 Hz equal 0.20 s.
Ask what would have to stay fixed for the comparison to be fair. Record those controls explicitly, then name the first uncontrolled condition that could reverse the conclusion. This is the practical transfer step, not an optional footnote.
List every field before attaching the word sag. This keeps the raw event recoverable if the classification rule changes.
Communication test: replace any sentence that says two things are “the same” or “different” with a sentence naming the operation, reference condition and supported consequence. The extra words are useful when they let a future reader repeat the decision without access to the original author.
Recalculate at 60 Hz with ten cycles. Duration becomes about 0.167 s, so cycle count and seconds must not be swapped.
Decision check: end with an action or interpretation that follows from the stated conditions, followed by a boundary sentence explaining when a specialist rule, local procedure or additional measurement is needed.
Real instruments can use overlapping windows, channel aggregation and hysteresis defined by their standard.
An interruption trace can retain a small nonzero residual voltage because of coupling, measurement noise or an alternate path. “Interruption” does not require the recorded sample to be mathematically zero.
Write the evidence in the order a reader can audit: the Utilities quantity, its reference object, the observation or browser rule, the transformation, and the decision being made. Keeping those fields visible prevents a familiar label from hiding a changed basis.
A second trace falls from 1.00 to 0.04 per unit for 1.4 s. Under the stipulated 0.10 per-unit boundary, record it as an interruption.
First-principles test: ask what physical, observational or algorithmic operation produces each value in this chapter. Then ask whether the second value could change while the first stays fixed. If it can, the two labels are not substitutes. For Utilities, this counterfactual is often clearer than a definition learned in isolation.
Raise the minimum to 0.12 per unit with all else fixed. It now sits above the stipulated interruption boundary and needs a different event label.
Evidence check: another reader should be able to reproduce the narrow result from the stored definitions, inputs and intermediate steps, then identify the nearest stronger claim that the evidence does not support.
Do not transfer one standard’s exact boundary into another report without checking its edition.
A three-phase supply may sag on one, two or three channels. Collapsing them into one minimum can hide whether the customer equipment saw unbalance, phase loss or a common event.
Trace the mechanism before interpreting the result. For Utilities, mark which value belongs to the source, which belongs to the measurement or layout algorithm, and which is a reported summary. A calculation is useful only when those roles remain attached.
Record phases A/B/C as 0.72, 0.94 and 0.95 per unit for the same 0.18 s interval.
Practical transfer: build two records that share the headline number but differ in the condition discussed here. If the correct interpretation changes, the omitted condition is decision-relevant and belongs in the report. This test turns a vocabulary distinction into a reusable Utilities checking habit.
Change only phase B to 0.08 per unit. The event receipt must show the channel-specific classifications rather than one site-wide word.
Practical check: keep one counterexample beside the worked case. If the wording fails when that one condition changes, narrow the sentence until it describes only the evidence actually in hand.
Phase-to-neutral and phase-to-phase measurements answer different questions.
The voltage trace classifies the observed event; it does not by itself prove the cause. Faults, motor starting, transformer energisation and network switching can produce different signatures and require corroborating evidence.
Use a compact decision table with columns for condition, input, operation, result and limit. This makes the comparison reproducible and exposes where a missing field would force an assumption instead of an observation.
A 0.70 per-unit trace coincides with a feeder protection log, but label the cause “consistent with” until timing and topology are reconciled.
Misconception check: a plausible output does not prove the right reference, denominator, ancestor, spectrum or observation rule was used. Reconstruct one intermediate step and one changed-context prediction. If either cannot be stated, keep the result provisional rather than strengthening its wording.
Remove the protection log while preserving the waveform. The event class stays supported and the causal claim weakens.
Transfer check: repeat the reasoning with one consequential condition changed while holding the rest fixed. State what changes, what stays comparable and why the original label alone was insufficient.
Causal diagnosis can require utility records, site wiring knowledge and specialist power-quality analysis.
RMS voltage summarises waveform energy over a window. Different window lengths and alignment can report different minima for the same rapidly changing waveform. Store method and raw waveform when available.
Preserve the raw inputs beside the derived result. If a later reader changes one condition, the original Utilities reasoning should still be recoverable without reverse-engineering a rounded headline or an unlabeled screenshot.
One-cycle windows report 0.70 per unit; a longer aggregated window reports 0.78 because recovery samples enter the average.
Evidence design: retain both a positive case and a near-miss. The positive case shows when the rule applies; the near-miss changes one consequential condition and shows when the conclusion fails. Together they teach the boundary of the Utilities mechanism better than another unlabelled example.
Shift the window by half a cycle and check whether the minimum changes while the physical waveform remains fixed.
Audit check: retain the source version, units, settings, exclusions and rounding rule. A future correction should be able to alter one field without rebuilding the entire explanation from memory.
A screenshot without sampling rate, window and aggregation cannot support a precise threshold decision.
An event monitor needs start and end thresholds. Hysteresis or a recovery threshold prevents repeated start-stop counts when voltage oscillates near the boundary.
Ask what would have to stay fixed for the comparison to be fair. Record those controls explicitly, then name the first uncontrolled condition that could reverse the conclusion. This is the practical transfer step, not an optional footnote.
Start the fictional sag below 0.90 per unit and end it only after voltage exceeds 0.92. A trace bouncing between 0.895 and 0.905 remains one event.
Communication test: replace any sentence that says two things are “the same” or “different” with a sentence naming the operation, reference condition and supported consequence. The extra words are useful when they let a future reader repeat the decision without access to the original author.
Remove hysteresis and count threshold crossings. The same physical disturbance fragments into several apparent events.
Decision check: end with an action or interpretation that follows from the stated conditions, followed by a boundary sentence explaining when a specialist rule, local procedure or additional measurement is needed.
Instrument defaults must be documented rather than assumed.
When phase events overlap in time, an aggregation rule decides whether to report one polyphase event or several channel events. Both the underlying channels and the aggregate should remain recoverable.
Write the evidence in the order a reader can audit: the Utilities quantity, its reference object, the observation or browser rule, the transformation, and the decision being made. Keeping those fields visible prevents a familiar label from hiding a changed basis.
Phase A sags from 0–0.20 s and phase B from 0.05–0.22 s. The site event spans 0–0.22 s under the stipulated aggregation.
First-principles test: ask what physical, observational or algorithmic operation produces each value in this chapter. Then ask whether the second value could change while the first stays fixed. If it can, the two labels are not substitutes. For Utilities, this counterfactual is often clearer than a definition learned in isolation.
Separate the intervals by 2 s. The aggregation rule may now create two events even when the cause is shared.
Evidence check: another reader should be able to reproduce the narrow result from the stored definitions, inputs and intermediate steps, then identify the nearest stronger claim that the evidence does not support.
Aggregation conventions affect counts and cannot be inferred from a single event total.
CHAPTER 12 OF 20 · Audit the evidence
12. Separate equipment response from supply label
A sag can reset one device while another rides through; an interruption can be bridged by stored energy. Equipment outcome is evidence about susceptibility and ride-through, not a redefinition of the supply event.
Trace the mechanism before interpreting the result. For Utilities, mark which value belongs to the source, which belongs to the measurement or layout algorithm, and which is a reported summary. A calculation is useful only when those roles remain attached.
The 0.70 per-unit, 0.20 s sag resets a contactor but a laptop continues on its battery.
Practical transfer: build two records that share the headline number but differ in the condition discussed here. If the correct interpretation changes, the omitted condition is decision-relevant and belongs in the report. This test turns a vocabulary distinction into a reusable Utilities checking habit.
Replace the contactor with a ride-through controller. The supply trace is unchanged while the process outcome improves.
Practical check: keep one counterexample beside the worked case. If the wording fails when that one condition changes, narrow the sentence until it describes only the evidence actually in hand.
Equipment immunity curves, controls and safety functions need device-specific assessment.
A large motor start may create a voltage depression without loss of upstream supply. Compare voltage depth, duration, current rise and timing to test the explanation.
Use a compact decision table with columns for condition, input, operation, result and limit. This makes the comparison reproducible and exposes where a missing field would force an assumption instead of an observation.
The monitor records 0.82 per unit for 0.8 s while current rises sixfold and then settles.
Misconception check: a plausible output does not prove the right reference, denominator, ancestor, spectrum or observation rule was used. Reconstruct one intermediate step and one changed-context prediction. If either cannot be stated, keep the result provisional rather than strengthening its wording.
Hold the voltage trace fixed but remove the current rise. The motor-start explanation loses its strongest corroboration.
Transfer check: repeat the reasoning with one consequential condition changed while holding the rest fixed. State what changes, what stays comparable and why the original label alone was insufficient.
Coincidence does not establish that the motor caused every downstream symptom.
A protection operation can create an interruption followed by restoration and possibly another event. Preserve the sequence instead of averaging the interval into one vague “power dip.”
Preserve the raw inputs beside the derived result. If a later reader changes one condition, the original Utilities reasoning should still be recoverable without reverse-engineering a rounded headline or an unlabeled screenshot.
Record 0.03 per unit for 0.5 s, recovery to 0.97 for 1.0 s, then 0.04 for 0.4 s.
Evidence design: retain both a positive case and a near-miss. The positive case shows when the rule applies; the near-miss changes one consequential condition and shows when the conclusion fails. Together they teach the boundary of the Utilities mechanism better than another unlabelled example.
Shorten the recovered interval below the monitor’s event-separation rule and see whether the report aggregates the sequence.
Audit check: retain the source version, units, settings, exclusions and rounding rule. A future correction should be able to alter one field without rebuilding the entire explanation from memory.
Protection settings and feeder topology are utility-specific.
CHAPTER 15 OF 20 · Transfer the reasoning
15. Relate events to continuity metrics carefully
SAIFI and SAIDI summarise sustained interruption frequency and duration across customers. A site power-quality monitor can detect sags and brief interruptions that do not enter those indices in the same way.
Ask what would have to stay fixed for the comparison to be fair. Record those controls explicitly, then name the first uncontrolled condition that could reverse the conclusion. This is the practical transfer step, not an optional footnote.
Keep a 0.20 s sag in the waveform ledger while leaving a sustained-interruption index unchanged.
Communication test: replace any sentence that says two things are “the same” or “different” with a sentence naming the operation, reference condition and supported consequence. The extra words are useful when they let a future reader repeat the decision without access to the original author.
Extend the loss to the applicable sustained threshold and identify which additional customer and restoration records are needed.
Decision check: end with an action or interpretation that follows from the stated conditions, followed by a boundary sentence explaining when a specialist rule, local procedure or additional measurement is needed.
Do not derive system reliability indices from one monitor trace.
Some analyses use declared nominal voltage; others may use a sliding or pre-event reference. A drifting reference can change the per-unit minimum without changing measured volts.
Write the evidence in the order a reader can audit: the Utilities quantity, its reference object, the observation or browser rule, the transformation, and the decision being made. Keeping those fields visible prevents a familiar label from hiding a changed basis.
At 161 V, use 230 V to obtain 0.70 per unit, then use a pre-event 220 V to obtain about 0.732.
First-principles test: ask what physical, observational or algorithmic operation produces each value in this chapter. Then ask whether the second value could change while the first stays fixed. If it can, the two labels are not substitutes. For Utilities, this counterfactual is often clearer than a definition learned in isolation.
Move the result near a boundary and show why the reference rule becomes decision-relevant.
Evidence check: another reader should be able to reproduce the narrow result from the stored definitions, inputs and intermediate steps, then identify the nearest stronger claim that the evidence does not support.
Reference selection must follow the adopted method rather than the desired label.
The everyday phrase “the power went out” can describe a visible reset, but a technical report should name the observed voltage event and the equipment response separately.
Trace the mechanism before interpreting the result. For Utilities, mark which value belongs to the source, which belongs to the measurement or layout algorithm, and which is a reported summary. A calculation is useful only when those roles remain attached.
Rewrite “a 200 ms outage shut the line” as “voltage retained 0.70 per unit for 200 ms; the contactor opened.”
Practical transfer: build two records that share the headline number but differ in the condition discussed here. If the correct interpretation changes, the omitted condition is decision-relevant and belongs in the report. This test turns a vocabulary distinction into a reusable Utilities checking habit.
Remove the waveform record. Keep the process symptom but mark the supply classification unverified.
Practical check: keep one counterexample beside the worked case. If the wording fails when that one condition changes, narrow the sentence until it describes only the evidence actually in hand.
Plain-language incident reports can remain readable while preserving evidence boundaries.
A defensible conclusion states reference voltage, retained RMS, duration, channels, method, threshold rule, equipment effects and causal uncertainty.
Use a compact decision table with columns for condition, input, operation, result and limit. This makes the comparison reproducible and exposes where a missing field would force an assumption instead of an observation.
Write: “On phase A, one-cycle RMS fell to 161 V, 0.70 pu of 230 V, for 0.20 s; this meets the stipulated sag rule, while cause remains under review.”
Misconception check: a plausible output does not prove the right reference, denominator, ancestor, spectrum or observation rule was used. Reconstruct one intermediate step and one changed-context prediction. If either cannot be stated, keep the result provisional rather than strengthening its wording.
Ask a second reader to reclassify the trace under a different declared threshold without rereading the waveform.
Transfer check: repeat the reasoning with one consequential condition changed while holding the rest fixed. State what changes, what stays comparable and why the original label alone was insufficient.
The report does not determine contractual responsibility or compliance.
Verify time synchronisation, channel labels, nominal voltage, probe range, event thresholds, firmware, calibration status and raw-data retention before trusting an event count.
Preserve the raw inputs beside the derived result. If a later reader changes one condition, the original Utilities reasoning should still be recoverable without reverse-engineering a rounded headline or an unlabeled screenshot.
Compare monitor time with a protection log and correct a 4.2 s clock offset before causal matching.
Evidence design: retain both a positive case and a near-miss. The positive case shows when the rule applies; the near-miss changes one consequential condition and shows when the conclusion fails. Together they teach the boundary of the Utilities mechanism better than another unlabelled example.
Leave the offset uncorrected and show how a plausible but wrong cause can be selected.
Audit check: retain the source version, units, settings, exclusions and rounding rule. A future correction should be able to alter one field without rebuilding the entire explanation from memory.
Installation and high-voltage work require qualified personnel and safe procedures.
CHAPTER 20 OF 20 · Report and continue
20. Continue through verified utilities routes
Service continuity versus adequacy concerns different system questions; SAIFI, SAIDI and CAIDI concern customer interruption indices; overlapping interruptions concern event aggregation. These routes complement rather than duplicate the voltage-event analysis.
Ask what would have to stay fixed for the comparison to be fair. Record those controls explicitly, then name the first uncontrolled condition that could reverse the conclusion. This is the practical transfer step, not an optional footnote.
Place the waveform event, customer interruption record and system index in three separate ledger rows.
Communication test: replace any sentence that says two things are “the same” or “different” with a sentence naming the operation, reference condition and supported consequence. The extra words are useful when they let a future reader repeat the decision without access to the original author.
Change the reader job from classifying one trace to comparing annual customer experience and choose the appropriate sibling.
Decision check: end with an action or interpretation that follows from the stated conditions, followed by a boundary sentence explaining when a specialist rule, local procedure or additional measurement is needed.
Each neighbouring article retains its own calculation point and limits.
Continue through the verified reading route
Service continuity and supply adequacy
Service continuity and supply adequacy. Use this when a reliable connection and sufficient system capacity answer different planning questions.
SAIFI, SAIDI and CAIDI
SAIFI, SAIDI and CAIDI. Use this when customer interruption counts and durations are aggregated into reliability indices.
Overlapping interruptions
Overlapping interruptions. Use this when several event records overlap in time or customers.
Frequently asked questions
Is every visible dimming a voltage sag?
No. Dimming is an equipment symptom; a classified sag needs a measured retained RMS voltage, duration and declared rule.
Must an interruption measure exactly zero volts?
No. A standard can classify a near-zero retained voltage below its threshold as an interruption.
Can a short sag affect equipment?
Yes. Susceptibility depends on depth, duration, phase, controls and ride-through, so the response is recorded separately.
Authoritative sources and further reading
IEEE 1159-2019. The official IEEE standard page identifies the recommended practice for monitoring electric power quality and its event terminology.
IEC 61000-4-30:2025. The official IEC page describes power-quality measurement methods, aggregation and uncertainty for relevant phenomena.
IEC 61000-4-11:2020. The official IEC page covers immunity tests for voltage dips, short interruptions and voltage variations in low-voltage equipment.
Return to the Utilities subject guide, The Gold Standard of Education, or the How X Works hub.