410,000 sq ft distribution center with fixture-level controls

The building

This is a representative scenario built from the kind of building we see most in North Mississippi: a big, busy, unglamorous box. The numbers are design-stage projections, not measured results.

The building is a 410,000 sq ft distribution center from the mid-2000s, with 32 ft clear height, about thirty aisles of selective racking, a bulk floor-storage area, 48 dock doors and a 9,000 sq ft office block. An earlier retrofit replaced metal halide with first-generation LED high-bays, roughly 640 of them at about 185 W each. Nobody added controls. A panel time clock turns everything on at 4:30 a.m. and off near midnight, six days a week, so storage aisles that see a forklift a few times an hour sit at full output for about 20 hours a day.

Connected lighting load is about 135 kW, and the lighting uses roughly 850 MWh a year.

Constraints

• The building never really stops. Two shifts run on weekdays and a partial shift on Saturday. Installation happens at night, section by section, with lifts working around live aisles.
• The existing LEDs aren’t old. Their drivers dim over 0-10V but have no auxiliary power for sensors, and every sensor retrofit still means a lift at 32 ft. We’d price both paths, adding sensors to the existing fixtures or replacing them with LLLC fixtures, and expect replacement to win once labor and incentives are counted.
• Pre-approval comes first. The building is served by a TVA local power company. TVA EnergyRight requires pre-approval before work starts and gives extra credit for networked controls, so nothing is ordered until written pre-approval arrives.
• Steel, stretch wrap and liquids are hard on 2.4 GHz radio, and inventory changes the RF picture every week.
• Forklift safety. No operator should ever drive into a dark aisle.

What we’d do, by phase

Survey (weeks 1–3). Walk every aisle on both shifts, inventory fixtures and circuits, take light readings at pick faces and pull twelve months of bills. Loggers go on a few circuits and a few aisles during the first visit, to show how often aisles are actually visited.

Bench (weeks 3–5). Put two candidate DLC-listed LLLC high-bays and their gateways on the bench in the Lytbit Lab. We’d check sensor coverage, group behavior, what the fixtures do when the gateway drops, and how closely their energy reporting matches a reference meter. Before proposing a pilot aisle, we’d ask program staff whether pilot fixtures would affect eligibility.

Design (weeks 5–9). Draw the Zoning Atlas: aisle groups, travel aisles, dock zones and the office. Write the sequences in the schedule above, set trim from measured pick-face illuminance rather than catalog output, and specify energy monitoring and C137.9-format configuration reports. The gateway gets its own VLAN or a cellular link, never the warehouse management network. Then we’d prepare the pre-approval package and coordinate submission with the local power company and EnergyRight program staff.

Oversee (about 10 weeks after pre-approval). The owner’s electrical contractor installs in eight night sections. We’d check submittals against the exact products in the application, because a substitution can cost eligibility, and walk each section before the lifts leave.

Commission (rolling). Test each section as it’s finished, so the night crew comes back to working aisles, then test the whole building: sequences, dock inputs, after-hours mode and trim. The configuration report and test logs feed the post-installation incentive paperwork.

Operate (12 months). Lytbit Ledger meters every fixture against the baseline. Ledger Watch flags sensors that never see vacancy, which usually means a bad aim or an aisle group that no longer matches the racking.

Projected outcome

Measure	Current operation	Projected
Connected lighting load	about 135 kW	about 100 kW
Annual lighting energy	about 850 MWh	270–360 MWh
Peak lighting demand	about 135 kW	about 88–101 kW
Full-output hours in storage aisles	about 20 h/day	5–8 h/day
Fault and energy data	none	per fixture

The low end assumes aisle traffic turns out heavier than the loggers suggest. Both ends sit close to DLC’s published figure, which estimates average savings of about 63% for NLC with LLLC. Incentive value depends on program rules at the time of application; we’d estimate it during design and confirm it at pre-approval.

What we’d watch out for

• Aisle-end sensors seeing cross-aisle traffic and lighting aisles nobody entered. We’d mask lenses and test every aisle end.
• Vacant levels set too low, too soon. Start at 20%, then step to 10% after two weeks without complaints. Otherwise supervisors override to full and the savings disappear.
• A purchase order placed before pre-approval, usually by someone trying to help the schedule.
• Racking changes. Aisle groups live in software, so regrouping is easy, but someone has to own it. We’d write that into the Sequence Book and the support plan.

What we’d do differently next time

Working through this scenario changed two of our defaults. First, loggers go up on the first site walk, not after it: real operating hours are the biggest uncertainty in the baseline, and every week of data narrows the range. Second, we’d commission the dock zones first rather than last. Docks are where people work fastest in mixed daylight, so that’s where complaints start, and a good first week there buys patience for the rest of the building.