Factory LED Retrofit: The Payback Math Buyers Check First

The scene: why the math comes before the catalog

A factory retrofit is an energy project wearing a lighting costume. The maintenance manager cares about lux and uptime, but the person releasing the budget cares about one sentence: what does it cost, what does it save, and when does it pay back. The vendors who lose these projects almost always lose them on arithmetic — a quote built from fixture price per unit, with no audit behind it, no maintenance line and no controls assumption. The vendors who win arrive with a measured baseline and a payback table the finance team can audit line by line.

The good news is that the arithmetic is genuinely simple, and industrial retrofits remain among the fastest-payback energy measures in any factory. Metal halide and T8-era installations burn 2-3 times the wattage of their LED equivalents for comparable delivered light — and that understates the case, because a 15-year-old high bay may be delivering 40-60 percent of its original lumens after reflector dirt and lamp decay, illustrative figures that a lux survey will confirm or correct. Below is the method in the order a buyer should run it, with a fully worked example whose numbers are illustrative throughout — real projects substitute their own tariffs, hours and quotes.

The five inputs that decide everything

Every credible payback model reduces to five measured or documented inputs. Collect them before requesting quotes, because each missing input becomes an optimistic assumption in someone else's favor:

Audit structure follows common energy-audit practice; the lux grid doubles as the acceptance criterion in the supply contract.

A worked example, line by line

Take a machining plant with 200 × 250 W metal halide high bays (285 W system draw each including ballast), running 6,000 hours a year on a two-shift pattern, at a blended tariff of $0.10 per kWh. Maintenance records show roughly half the lamps replaced annually at $90 per intervention and a steady trickle of ballast failures. The proposed retrofit: 100 W LED high bays with photometric confirmation, plus aisle presence sensors. Every number from here is illustrative — the method is what transfers.

All figures illustrative. Real projects substitute measured hours, the actual blended tariff and quoted equipment prices.

Against that, the illustrative project cost: 200 LED high bays at $95 each is $19,000, installation at $45 per point is $9,000, and sensors and commissioning add $4,000 — $32,000 total. Payback lands at roughly 11 months, with a five-year net of about $138,000 and a first-year return above 100 percent. Halve the operating hours to 3,000 and payback stretches toward two years — which is exactly why the hours line in the audit table is the one to measure rather than estimate. Add utility rebates or DLC-listed product incentives where they exist (our DLC guide covers the North-American case) and the equation shortens again. Numbers in this band are why lighting is usually the first measure approved in an energy program rather than the last.

Controls and sensors: the multiplier

Fixture-for-fixture swapping captures the wattage saving and stops there. The controls layer captures the hours, and in the right building it is worth a fifth to a third of the lighting bill on its own — illustrative range, typical of halls with intermittent occupancy. Three measures dominate: aisle-by-aisle presence sensing, which drops storage rows to a background level between forklift passes; daylight harvesting along clerestories and skylight bands; and high/low task tuning, which trims over-lit zones to the EN 12464-1 target rather than a commissioned maximum. Our smart lighting program covers the control stack, and the layout logic pairs with the warehouse design guide — the sensor zoning is drawn on the same plan as the fixtures, at the same time, or it will not get drawn at all.

Sequence the controls with the fixtures, not after them. Sensors commissioned before the racks are restocked or the machines reinstalled read an empty hall and tune themselves wrong; the trial aisle exists precisely to commission one complete operating cycle — occupied, empty, overnight — before the model is confirmed for the rest of the plant. Log the trial week at the distribution board and the before-and-after numbers write the acceptance report themselves.

Common mistakes in retrofit arithmetic

• Quoting like-for-like wattage. "Replace 250 W with 250 W" ignores optics and delivered lumens. The correct comparison is photometric: layout files before and after, at the work plane.
• Ignoring lumen decay in the baseline. A 15-year-old installation may deliver half its designed lux. The retrofit's "after" survey against a measured "before" grid is the only honest proof of improvement.
• Omitting maintenance entirely. The lamp-and-lift line is real money; leaving it out understates savings and hands the project a weaker case than the facts support.
• Buying the cheapest drivers. A fixture that fails at month 18 erases its entire saving in one service call. Driver brand and warranty terms belong in the financial model, not just the datasheet.
• Forgetting the lux floor. The goal is compliant, uniform light at minimum wattage — not maximum lumens per dollar of electricity. Over-lighting is a smaller waste than under-lighting a work station, but both fail the audit.
• No acceptance criteria. Without the before-grid, the after-grid and a uniformity threshold written into the contract, "done" becomes a matter of opinion.

Specifying and delivery: what to close with the factory

The sourcing checklist for a retrofit compresses into four documents. First, photometric files and a layout — IES/LDT per model, run against the measured plan, showing the after-lux grid and uniformity. Second, the BOM by name: LED package and binning window, driver brand, surge rating of 4-10 kV, ambient range for unheated halls and hot roof zones alike. Third, compliance papers: CE for the EU, UKCA for the UK, SAA for Australia, SASO/SABER for the Gulf — issued via certified partner factories and verifiable by certificate number, per our certification guide. Fourth, an acceptance protocol: the after-survey against the contract grid, plus spare-parts ratio (2-3 percent of drivers and optics) delivered with the goods.

Commercially, retrofit projects reward staged terms: samples of 1-5 pcs in 7-14 days, a one-aisle trial zone installed and surveyed before mass release, then bulk production 25-40 days after deposit — illustrative figures, confirmed per order. Fixture choice between formats is covered in the UFO vs linear comparison; the wider program, including the halls and ancillary zones that usually travel in the same container, sits with our warehouse and industry solutions team and the industrial lighting product line. One last sourcing note: ask for the warranty document itself, not the word — what the 3-5 year term covers, who pays freight on a claim, and how failures are proven, since those three clauses decide whether the paper warranty is worth the ink.