Voltage Sags Only During Heavy Summer Load: POCO vs Service Decision Tree

Why this matters

Lights dim and motors strain on hot afternoons when air conditioning across a neighborhood peaks, and the customer cannot tell whether the problem is on their side of the meter or the utility's. The stakes are real: a sag caused by a loose service-entrance neutral or a degraded line connection on the customer's side is a fire and equipment-damage hazard the contractor owns, while a sag caused by undersized utility transformer or distribution loading is the power company's to fix. Calling the wrong party wastes a truck roll and leaves either a hazard or a customer paying for utility-side trouble. The diagnostic discriminator is where the voltage drop appears: utility-side sags depress both legs roughly equally at the line side of the meter, while service-side problems show up as a drop that develops across a specific connection, a leg imbalance, or a neutral that floats. This tree localizes the drop to one side of the meter using readings any service truck can take.

Symptom presentation

The complaint is load-correlated and time-correlated: brownout symptoms appear during the hottest hours, ease at night, and worsen year over year as neighborhood load grows. Measured line-to-line and line-to-neutral voltage falls below nominal under load and recovers when large loads cycle off. A clamp on the service conductors shows current near or above the service rating during the sag window. Whether the two legs sag together or one leg sags more than the other is the first branch of the tree.

Quick checks

• Log voltage at the main lugs (line side of the main breaker) line-to-line and each leg line-to-neutral during the sag, and again at night. A voltmeter with min/max recording or a power-quality logger over 24 to 48 hours is the right tool.
• Compare measured nominal: 120/240 single-phase should hold within roughly plus or minus 5 percent at the service per ANSI C84.1 Range A.
• Clamp each service conductor and the neutral during the sag. Note total current versus service rating.
• Inspect the meter base, service-entrance neutral, and main lugs for heat discoloration, corrosion, or a loose lay-in lug.

Isolation tree

1. Read both legs to neutral at the main lugs during the sag. If both legs sag by a similar percentage and the neutral-to-ground voltage stays near zero, the depression is arriving from the utility: proceed to step 2. If one leg sags noticeably more than the other, or neutral-to-ground voltage rises during the sag, the problem is on the service side: proceed to step 4.

2. Take the same reading at the line side of the meter (utility connection point) and compare with the main lugs. Equal voltage at both means no drop across the service entrance; the depressed voltage originates upstream of the meter. Document and refer to the POCO.

3. Before referring, confirm the service-entrance connections are not adding to the drop: measure voltage drop from the weatherhead/line side to the main lugs under load. A drop that exceeds a few volts across that span points to a service-side connection even when both legs sag, so this measurement protects against blaming the utility for a contractor-fixable lug.

4. Service-side localization. A larger sag on one leg points to a high-resistance connection on that leg: walk voltage drop across the meter jaws, the main lug, and each splice on the affected leg under load. A rising neutral-to-ground voltage points to a loose or corroded service-entrance neutral, which is both a sag source and a serious hazard.

Reading the data correctly

A few measurement habits prevent false conclusions. Take simultaneous readings: a single roving meter cannot compare meter-line-side and main-lug voltage during the same instant of a transient sag, so use two loggers or a logger plus a witnessed spot reading at a known load condition. Record under a repeatable load (have the customer run the AC plus a couple of large loads) rather than chasing random fluctuations. Note that ANSI C84.1 Range A allows roughly plus or minus 5 percent at the service point; a sag to 114 V on a 120 V nominal leg is at the edge of acceptable utility delivery and may legitimately belong to the POCO even though the customer perceives it as a fault. Distinguish a sag (depression under load that recovers) from a sustained low voltage (always low), because the latter points more toward a chronically undersized service or feeder than toward a connection that degrades only under peak current.

Confirming diagnosis

Confirm a utility-side sag by logging voltage at the meter line side over the peak window and showing the depression is already present there with no added drop across the service entrance; hand the POCO the timestamped log. Confirm a service-side connection by measuring a localized voltage drop across one specific joint that disappears when the load drops, then re-terminating and re-measuring near zero drop. Confirm a neutral fault by the neutral-to-ground voltage collapsing to near zero after re-terminating the service neutral. Confirm a genuinely undersized service by an Article 220 calculation showing demand near or above the service rating with all connections proven sound, which moves the case to the service-upgrade path rather than a repair.

Remediation

• Utility-side sag (undersized transformer/distribution): document and refer to the POCO with the logged data.
• High-resistance service connection: re-terminate meter lugs, main lugs, or splices to torque spec; replace corroded hardware.
• Loose/corroded service neutral: repair or replace the neutral connection, coordinate with the utility where the connection is on their side of the demarcation.
• Genuinely undersized customer service for present load: see the service-upgrade decision path.

References

• ANSI C84.1, Electric Power Systems and Equipment Voltage Ratings (Range A/Range B)
• NFPA 70 (NEC) Article 215, Feeders (voltage-drop guidance)
• NFPA 70 (NEC) Article 230, Services
• NFPA 70 (NEC) Article 250.24, Grounding Service-Supplied Systems (service neutral)
• IEEE 1159, Recommended Practice for Monitoring Electric Power Quality