communityfix.org

Write noise limits in octave bands and dBC, measured at the residential property line, as a permit condition

#00215

Replace single-number dB(A) ordinances with limits that also constrain C-weighted and per-octave-band levels at the nearest residence, set as a permit condition rather than a complaint threshold. Prince William County proposed exactly this after observing a 90 Hz tone at a specif

Parent issue

#00205 Continuous low-frequency noise from cooling plant reaches homes at levels ordinances are not written to measure

Location

city

Description

The proposal

A local authority writes three things into the ordinance and then into each permit: a C-weighted limit alongside the A-weighted one, per-octave-band limits covering the 31.5–250 Hz range where cooling plant tones sit, and a measurement point at the nearest residential property line rather than the facility boundary. Compliance is a continuing permit condition with a defined remedy, not a threshold that triggers a complaint process after the fact.

Why it would work

The mechanism of harm is tonality and persistence, not average loudness. A-weighting was derived from equal-loudness contours for quiet broadband sound and systematically discounts low frequencies, so a facility can pass a 55 dB(A) test while producing a hum that is clearly audible indoors at night. Adding dBC captures low-frequency energy; octave-band limits catch a discrete tone that broadband metrics average away.

This is not speculative drafting. Prince William County's March 2025 proposal adds exactly these measures, stating that "the A-weighted scale doesn't effectively address the low-frequency noise generated by Data Centers", citing an observed 90 Hz tone at a specific facility, and proposing residential intermittent limits of 48 dBA day and 43 dBA night (Prince William County).

Two ordinance defects should be fixed at the same time: exemptions for air conditioning equipment (which some ordinances apply to industrial chiller arrays), and the general design assumption that noise complaints concern intermittent events rather than continuous plant.

Implementation path

This requires no new state or national authority. A county or municipality commissions a baseline acoustic survey establishing existing ambient levels at representative receptors, adopts limits expressed as a permitted increase over that measured baseline plus absolute caps, and attaches them to each special exception. The Lyver study is a usable model for the survey stage, having measured 25 receptor locations and modelled buildout (Lyver). Removing by-right development—as Loudoun County did in March 2025—is what makes permit conditions attachable at all (Holland and Knight).

Setbacks are the complement, not the substitute. Sound attenuates with distance regardless of spectrum, so minimum separation from residential zoning does work that no metric alone can.

Trade-offs and limitations

  • Prospective only. Existing facilities were permitted under old rules and cannot generally be brought under new conditions without a nuisance action or negotiated retrofit.
  • Enforcement capacity. Octave-band measurement requires equipment and trained staff that small jurisdictions typically lack; a limit nobody measures is not a limit.
  • Evidence gap. The specific numbers chosen cannot currently cite a dose-response study for this source type. The engineering rationale for constraining tonality is nonetheless independent of the epidemiology.
  • Cost falls on the operator in equipment selection and enclosure design—the intended effect, but a real cost that should be stated rather than assumed away.

Sub-issues

0
View all
No sub-issues yet. Add the first one →

Case studies

2
View all

communityfix.org