Key takeaways
·Luminaire efficacy at the terminals (lm/W) is the only number that survives a tender review — chip efficacy such as “up to 200 lm/W” is a component claim, not a product claim.
·2026 mainstream export-grade ranges: linear 140–170 lm/W, UFO high bay 140–165 lm/W, panel 125–150 lm/W, deep recessed downlight 110–140 lm/W.
·Lifetime must be quoted with case temperature. L90 at 25 °C and L80 at 50 °C describe two different products, even with the same 50,000-hour headline.
·Glare and flicker reject more fixtures from a real project than efficacy does: UGR ≤ 19 indoors, modulation below 8 % or SVM below 1 anywhere near a camera.
·In solar systems every watt of load is paid for twice — once at the LED, once at the panel and battery — so a 10 % efficacy gain cuts system cost directly.
Why the wattage conversation is over
For most of the last decade, LED procurement was a wattage negotiation. A 150 W high-pressure sodium fixture became a "60 W LED equivalent", and everybody moved on. That shortcut no longer works, for three reasons:
·Utility rebate programmes and green building codes now specify delivered lumens per watt at the luminaire level, not the LED chip level.
·Cheap 2835-based fixtures can hit 130 lm/W on a test bench and deliver 95 lm/W in a factory at 45 °C ambient.
·Solar systems make the stakes higher: every watt of load must be paid for twice — once at the LED, once at the panel and battery.
The three numbers that decide a project
1. Luminaire efficacy (lm/W). Ask for the LM-79 report of the finished fixture. If a supplier quotes chip efficacy ("up to 200 lm/W"), that is a component specification, not a product specification.
2. Thermal-end lifetime (L90 / L70). A 50,000-hour rating means nothing without the case temperature it was measured at. L90 at 25 °C and L80 at 50 °C describe very different products.
3. Glare and flicker. UGR for indoor, unified glare ratio for outdoor, and flicker percentage (or SVM) for anything near a camera. A high-efficacy fixture that flickers at 20 % is unusable in an office or a broadcast venue.
2026 reference specification table
The values below are realistic mid-market ranges for mainstream export-grade products — useful as a benchmark when you shortlist quotations.
Category | Typical system efficacy | CCT range | CRI | IP / IK | Control options | Typical warranty |
LED panel light (600×600) | 125–150 lm/W | 3000–6500 K | 80 / 90 | IP20–IP44 / IK05 | 0-10 V, DALI, sensor-ready | 3–5 years |
140–170 lm/W | 3000–5000 K | 80 / 90 | IP20–IP54 / IK07 | DALI-2, D4i, emergency module | 5 years | |
LED downlight (deep recess) | 110–140 lm/W | 2700–4000 K | 80 / 90 | IP44–IP65 / IK06 | Triac, DALI | 3–5 years |
UFO high bay | 140–165 lm/W | 4000–5700 K | 70 / 80 | IP65 / IK08 | Microwave sensor, 1-10 V | 5 years |
130–155 lm/W | 3000–5700 K | 70 / 80 | IP66 / IK08 | Photocell, DMX | 5 years | |
150–180 lm/W | 3000–5000 K | 70 / 80 | IP66 / IK08 | NEMA 7-pin, Zhaga, 10 kV SPD | 5–7 years | |
All-in-one solar street light | 160–200 lm/W (LED side) | 4000–6500 K | 70 / 80 | IP65 / IK08 | PIR + timer dimming | 3–5 years |
Decorative / architectural | 100–130 lm/W | 2200–4000 K | 90 | IP20–IP67 | DMX512, 0-10 V | 2–3 years |
Two notes on reading this table. First, solar fixtures show the highest LED-side efficacy because the panels and battery impose such a hard energy budget; the system number that matters is lumens per night per watt-hour of storage. Second, decorative and architectural products trade efficacy for colour quality and form factor — a CRI 95 wall washer at 110 lm/W is a better product than a CRI 80 one at 140 lm/W.
Where the lumens actually disappear
A 160 lm/W chip inside a 110 lm/W fixture lost 30 % of its output on the way to the floor. The usual distribution:
·Driver conversion: 8–12 % (a premium constant-current driver runs 90–93 %; a budget one 82–86 %).
·Optical losses: 8–15 % depending on diffuser type. A micro-prismatic PMMA panel loses less than an opal PC sheet, but costs more.
·Thermal derating: 5–12 % once junction temperature rises 40 K above the 25 °C test condition.
·Depreciation and dirt: 10–20 % over five years in an industrial environment without maintenance.
Any quotation that shows only the first number is hiding the other three.
Five questions to ask before you sign an RFQ
1.Can you send the LM-79 report for the finished fixture, with the model number on it?
2.What driver brand and model is inside, and what is its measured efficiency at full load?
3.At what ambient temperature and case temperature was the lifetime rating measured?
4.What is the flicker percentage at 50 % dimming, and does the driver meet IEEE 1789 "no risk" limits?
5.For solar products: what is the autonomy in days at the specified lighting hours, and what is the usable battery capacity at end of life?
Suppliers who answer all five in writing are usually the ones who also survive the installation.
A worked total-cost-of-ownership comparison
Take 100 fixtures running 4,000 hours a year at USD 0.12/kWh.
Item | 150 W HPS | 60 W LED (150 lm/W class) |
Connected load | 15.0 kW | 6.0 kW |
Annual energy | 60,000 kWh | 24,000 kWh |
Annual energy cost | USD 7,200 | USD 2,880 |
Lamp / driver replacement (5 yr) | USD 4,500 | USD 400 |
Five-year energy saving | — | USD 21,600 |
The fixture price difference is typically recovered in 12–20 months on a 4,000-hour schedule. Shorter operating hours push payback out; 24/7 industrial sites pull it in under a year.
Frequently asked questions
Is 200 lm/W realistic for a finished luminaire in 2026?
For high-end street and high bay fixtures using efficient mid-power LEDs and a well-managed thermal path, 180–200 lm/W is achievable at CRI 70. At CRI 90 or with a wide diffuser, expect 130–160 lm/W.
Do solar street lights need a higher efficacy LED than grid-connected ones?
Yes. Because panel and battery capacity dominate the bill of materials, every 10 % gain in LED efficacy can reduce the panel and battery by roughly 10 %, which is often the cheapest way to hit a target price.
Does higher efficacy always mean lower running cost?
Only if the optic and thermal design are sound. A narrow, efficient fixture aimed badly wastes more light than a modest one with correct photometry.
Choosing a supplier who can prove the numbers
LFL Lighting builds indoor, outdoor and solar luminaires to a single specification discipline: LM-79 data for the finished fixture, named driver brands, and a thermal report for every platform. If you are preparing a 2026 tender or refreshing a product line, send us your target efficacy, CCT and IP rating — we will return a comparable spec table and a quotation within two working days.
Contact: shally@longingforlight.com | Request the 2026 specification workbook
Related reading: Industrial & Commercial LED Lighting Directory (high bays, flood lights, tri-proof) · Comprehensive Commercial Indoor Lighting Analysis (panels, linear, downlights) · Next-Generation Solar Outdoor Lighting Systems.