光效是什么
Two floodlights both labelled "200 W" can produce 18,000 lumens and 32,000 lumens respectively. Wattage tells you what you pay; lumens tell you what you get; efficacy is the exchange rate between them. When you define luminous efficacy targets in a specification, you are directly fixing the energy cost of every hour the installation will ever run.
To define luminous efficacy rigorously, you need both halves of the fraction:
Luminous efficacy η = Φv (lm) ÷ P (W)
Luminous flux Φv is radiometric power weighted by the human eye's photopic sensitivity V(λ), as standardized by the CIE and catalogued in CIE S 017:2020. Because the eye peaks at 555 nm, one watt of perfect 555 nm radiation equals 683 lm — the absolute physical ceiling. White light, which must spread energy across the visible spectrum, tops out around 250–370 lm/W depending on CRI and CCT targets.
Every real light source loses energy at several stages, which is why practical values sit far below the ceiling. For a typical LED luminaire the chain looks like this:
The most common specification error is comparing a competitor's system (luminaire) efficacy against a vendor's source (chip) efficacy. A "200 lm/W LED" claim usually refers to the bare package at 25 °C and low drive current. After driver, thermal, and optical losses, the finished luminaire may deliver 140 lm/W. When you define luminous efficacy in tender documents, always specify luminaire efficacy, measured per IES LM-79-19, at stabilized operating temperature, including driver consumption. That single sentence eliminates the 20–40% inflation between chip marketing and delivered performance.
Efficacy competes with color quality. Raising CRI from 80 to 90 typically costs 10–15% efficacy, because high-CRI phosphor blends push more energy into the deep-red region where V(λ) sensitivity is low. Similarly, warm 2700 K sources test a few percent below cool 5000 K equivalents. This is why credible specifications define luminous efficacy together with CRI and CCT — "≥ 150 lm/W" alone invites bids with CRI 70 cool-white product that wins the efficacy race but fails the application.
| Technology | Typical Efficacy (lm/W) | Notes |
|---|---|---|
| LED luminaire (premium, 2024–2026) | 180–220 | System level; CRI 70–80, 4000–5000 K |
| LED luminaire (standard commercial) | 100–160 | System level, incl. driver and optics |
| LED package (chip, lab conditions) | 200–240+ | 25 °C, low drive current — not achievable in fixtures |
| Fluorescent T5 (incl. ballast) | 80–105 | Being phased out under RoHS/Ecodesign mercury rules |
| Metal halide | 65–115 | Rapid lumen depreciation lowers maintained efficacy |
| High-pressure sodium | 80–150 | High efficacy but CRI ~25 |
| Halogen | 15–25 | Banned for general service in EU since 2018 |
| Incandescent | 10–17 | ~95% of input becomes heat |
| Requirement / Benchmark | Threshold | Reference |
|---|---|---|
| EU minimum for mains LED light sources | Efficacy floor via Pmax formula (≈ 85–120 lm/W depending on flux and CRI) | EU Regulation 2019/2020, Annex II |
| DLC Premium listed luminaires (typical categories) | ≥ 125–150 lm/W depending on product class | DLC Technical Requirements V5.1 |
| China LED downlight/luminaire energy grades | Grade thresholds from ~65 to ≥ 105 lm/W by class | GB 30255-2019 |
| Measurement method (photometry + electrical) | Absolute photometry, stabilized conditions | IES LM-79-19 |
| Lumen maintenance projection | L70/L80/L90 lifetime from 6,000 h+ test data | IES LM-80-15 + TM-21-19 |
| Vocabulary / formal definition | "Luminous efficacy of a source", unit lm/W | CIE S 017:2020 (ILV) |
Efficacy translates most directly into money where fixtures are numerous, powerful, and long-burning. Replacing 400 W metal halide high-bays (~75 lm/W maintained) with 150 lm/W LED units cuts connected load by roughly 60% for equal illuminance. At scale the arithmetic is decisive: 300 fixtures × 12 h/day, moving from 100 lm/W to 160 lm/W product at equal lumen output, saves ~37% of lighting energy — often six figures over a 10-year life. For these projects, define luminous efficacy minimums at 150–170 lm/W system level, CRI ≥ 70, verified by LM-79 reports.
Offices trade a little efficacy for visual comfort: UGR ≤ 19 optics (micro-prismatic diffusers) absorb 8–15% of flux, so realistic targets are 110–140 lm/W for compliant panels versus 160+ for bare-lens industrial product. Energy codes drive the numbers indirectly through Lighting Power Density (LPD, W/m²) limits — ASHRAE 90.1-2022 allows roughly 6.5 W/m² for open offices, only achievable with high-efficacy sources plus controls. Specify efficacy and UGR together; either alone produces a bad building.
Municipal tenders now routinely demand ≥ 130–150 lm/W plus photometric distribution requirements (EN 13201 / IES RP-8-22 classes). Efficacy matters doubly here because street lights burn 4,000+ hours a year and the optics (Type II/III distributions, full cutoff) consume part of the flux budget. Maintained efficacy is the honest metric: require TM-21 projections (L90 ≥ 100,000 h is common for quality product) so year-10 performance, not day-1 performance, meets the lux requirement.
Here the priority order reverses: CRI ≥ 90 (with R9 ≥ 50) and precise beam control outrank raw lm/W, and accepting 100–130 lm/W is a rational trade. The professional approach is still to define luminous efficacy floors — just calibrated to the color-quality class — so vendors cannot quietly substitute low-efficacy COBs while meeting the CRI line.
Plant lighting abandons lumens entirely, because plants do not share the human V(λ) curve. The equivalent metric is photon efficacy in µmol/J (PPE), measured to the same LM-79 electrical rigor. If a horticultural supplier quotes lm/W, that is a red flag — the industries define efficacy differently for good physical reasons.
Browse 20,000+ lighting products with certified LM-79 specifications, compare efficacy, CRI and prices, and connect with verified manufacturers on Compare2Best.
Browse Lighting Products →