Sound Attenuation and Setback Standard

Purpose

A generator that seems quiet enough standing next to it during a daytime demo can still trip a nighttime ordinance limit at the property line the first time it exercises at 2 AM, and by then the pad is poured and the fix is a retrofit enclosure or a full relocation instead of a siting decision made before anything was placed. This procedure sets the standard for computing, not guessing, the sound level a specific unit will produce at a specific neighbor's window and property line, comparing it against the actual local limit, and confirming the estimate with a measured reading once the unit runs under real load.

Scope

Covers sound-level calculation, ordinance and covenant comparison, mitigation selection, and post-install measurement for a residential or light commercial standby unit's siting. It does not cover exhaust termination siting or CO safety, owned by the exhaust clearance and CO safety standard; pad dimensioning and anchoring, owned by the concrete pad and mounting standard; or state-specific LP tank setback distances, owned by the relevant jurisdiction-specific reference where one exists for that state. This procedure is the general method; a jurisdiction-specific reference, where one exists, supplies the actual adopted number for that state.

Roles and responsibilities

Procedure

1. Pull the manufacturer's sound rating with its test distance and standard, and do not compare ratings measured differently. Get the rated dBA figure from the spec sheet along with the distance and test standard it was measured at (commonly a stated distance under a standard such as ISO 8528-10). Acceptance: a sound figure captured in writing with its measurement distance and standard noted, not the number alone. Wrong looks like comparing one unit's rating at one distance against a competing unit's rating at a different distance as if they were equivalent. Stop rule: no siting calculation proceeds on a sound number whose measurement distance is unknown; pull it from the actual spec sheet. Hazard: none; a records step.

2. Measure the actual on-site distance to the property line and the nearest neighboring window, then compute the estimated level using the distance-doubling method. Measure both distances on site, then apply the rule that sound pressure drops approximately 6 dB each time distance from the source doubles in an open, unobstructed field, noting that a solid barrier between source and receiver can add attenuation while a reflective wall can reduce it. Acceptance: a computed dB figure for both the property line and the nearest window, each derived from a measured on-site distance, not assumed. Wrong looks like eyeballing "it's pretty far, should be fine" without running the doubling calculation from an actual measured distance. Stop rule: no siting sign-off without the calculation on paper for both distances. Hazard: none; a measurement and calculation step.

3. Compare the computed figure against the applicable local noise ordinance, HOA covenant, and code minimum clearance, confirming the current adopted limit rather than assuming a typical one. Check the specific jurisdiction's ordinance for its daytime and nighttime limits at the property line, and the HOA covenant where one applies, since these vary block to block in some counties and change with adopted-code updates. Acceptance: a written pass or fail against each applicable limit, sourced by ordinance section or HOA document. Wrong looks like assuming "a typical nighttime limit" without checking the specific jurisdiction's current figure. Stop rule: any computed figure exceeding an applicable limit blocks the current siting plan and routes to step 4's mitigation selection before anything is poured. Hazard: none; a comparison step.

4. Where the computed level exceeds a limit, select and size a mitigation, then recompute against the same limits. Choose among repositioning (doubling distance reduces the estimate by roughly 6 dB, per the same method as step 2), a sound-attenuating enclosure (use the manufacturer's own rated reduction only when tested at a matching reference distance, not stacked onto a raw comparison), or a barrier (partial line-of-sight blocking with no guaranteed dB reduction unless the barrier itself carries a rated attenuation value). Acceptance: a revised computed figure after mitigation, re-checked against every limit from step 3. Wrong looks like buying an accessory because a supplier says it helps, without re-running the calculation to confirm it actually closes the gap. Stop rule: no mitigation is accepted as sufficient without a recomputed figure that passes. Hazard: fastening an enclosure panel near the control compartment on an energized unit risks contact with live low-voltage control wiring; de-energize control power per the manufacturer's service procedure before working inside the compartment.

5. Verify the same site plan simultaneously satisfies physical setback minimums independent of sound. Check the chosen location against property-line and combustible-clearance minimums and, for a propane install, tank setback distance, since a location chosen purely to minimize dB can violate a setback minimum nobody re-checked. Acceptance: one site plan checked against sound compliance, code setback minimums, and HOA covenant simultaneously, with no conflict remaining. Wrong looks like solving a sound problem by moving the unit closer to the property line, which passes step 3 again but now fails a setback minimum. Stop rule: any move triggered by step 4 re-runs step 3 against the new distance before the plan is final. Hazard: none; a cross-check against the other siting SOPs.

6. Commission with an actual sound-level meter reading at the property line and nearest window during a real load step, and compare it to the step 2 estimate. Take the reading with the unit under load, not idle, since idle readings run well below loaded operation. Acceptance: measured dBA at load meets the applicable limit, recorded with the meter reading, time, and load percentage, and compared against the step 2 estimate to confirm the method held. Wrong looks like measuring at no-load idle and signing off on a number the unit will never actually produce during a real outage. Stop rule: any measured exceedance stops sign-off and returns to step 4's mitigation menu, now driven by a measured number instead of the calculated one. Hazard: taking a reading near a running, loaded unit puts the tech close to a moving fan, belt, and hot exhaust; keep clear of moving components and the exhaust discharge path while reading the meter.

7. Brief the customer on exercise-schedule noise timing and document the re-verification trigger. Confirm the auto-exercise schedule's timing against the ordinance's quiet hours, get written acknowledgment, and tell the customer that any future structural change near the unit (new window, enclosed porch, removed fence) requires the survey to be re-run. Acceptance: written acknowledgment on file, re-verification notice given. Wrong looks like scheduling the weekly self-test during posted quiet hours because nobody cross-checked the two. Stop rule: the job does not close without the acknowledgment on file. Hazard: none; a documentation step.

The record this produces

One sound and setback record per install. Fields: manufacturer rating with distance/standard, measured on-site distances, computed estimate at property line and nearest window, applicable ordinance/HOA figures and their source, mitigation selected and its recomputed figure, setback cross-check result, the measured commissioning reading with load percentage, and the customer's signed acknowledgment. The office reads it if a noise complaint is ever filed, since it is the shop's evidence the siting was calculated, not guessed.

One pass, with a step that failed

A 20 kW natural gas unit sited on a narrow suburban lot, property line 25 feet from the planned location. Step 1 pulled the manufacturer's rating: say the spec sheet lists 68 dBA at 25 ft under ISO 8528-10, an illustrative figure for this walkthrough. Step 2 measured the property line at that same 25 ft, so no distance reduction applies yet: estimated level at the property line, 68 dBA.

Step 3 is where the pass failed. The office pulled the jurisdiction's actual posted nighttime limit, 60 dBA at the property line, not assumed. At 68 dBA against a 60 dBA limit, the site exceeded it by 8 dB, and the stop rule blocked the plan before the pad was poured.

Step 4 worked the mitigation menu. Repositioning to 50 ft from the property line, one full doubling from the rated 25 ft, cut the estimate by 6 dB to 62 dBA under the same method, still 2 dB over the limit on its own. Adding the manufacturer's sound-attenuating enclosure, rated for a 5 dB reduction at the same 25 ft reference distance the base rating used, brought the estimate to 57 dBA, 3 dB under the 60 dBA limit. Step 5 re-checked the 50-ft location against combustible-clearance and property-line minimums (no tank setback applied on natural gas); both cleared at the new distance. Step 6 commissioned under a load-bank step: the measured reading at the property line came in at 58 dBA, close to the step 4 estimate and still under the 60 dBA limit, confirming the doubling-plus-enclosure method held. Step 7 briefed the customer, confirmed the weekly self-test was scheduled outside posted quiet hours, and filed the signed acknowledgment naming the 50-ft siting distance as the baseline for any future re-verification.

When it does not fit the usual case

The site has no enforced local noise ordinance but does have an HOA covenant: treat the covenant as the controlling limit in step 3 rather than skipping the comparison because no municipal ordinance exists. The unit is propane-fed and the LP tank's own setback conflicts with the sound-optimal location: resolve both simultaneously in step 5 rather than treating tank setback as a separate, later problem, and consult the state-specific LP setback reference for that jurisdiction's actual distance table. A neighbor complaint arrives after commissioning despite a passing measured reading: re-run step 6 at the actual complaint time and load condition rather than assuming the original reading still applies, since a different load step or wind condition can produce a different result.

References

• Exhaust clearance and CO safety standard, for exhaust siting and CO commissioning, a separate concern from sound.
• Concrete pad and mounting standard, for pad dimensioning this procedure does not duplicate.
• Generator codes and permits reference, for the general permit and placement-distance framework.
• Florida LP tank setback and permit requirements for standby generator installs, as an example of a jurisdiction-specific setback reference to consult for the actual adopted distance table where one exists for that state.
• NFPA 37 Standard for the Installation and Use of Stationary Combustion Engines and Gas Turbines, current adopted edition.
• ISO 8528-10, for generator sound-level measurement methodology and reference distances.