We recently had this question sent to us via the Hadco web site. I thought the response should be posted and shared to cut through some of the hype and get to some basic facts:
As engineers, we need some very specific decumentation regarding all of the wild efficiency claims for LED lighting for exteriors — from what we can tell, NO led is more efficient on lumens delivered to he pavement per watt consumed than the better Metal halide systems - we need this data FIRST, before we then factor in the lumen depreciation of MH, the lamp life issues, etc etc —- most vendors roll every thing into one over-simpllified blanket statment - -misleading owners, architects, etc - then lighting consultants have to dis prove the negative — please help us out here !!
Our response:
The explanation you have requested is somewhat lengthy and deals specifically with the inherit design issues surrounding todays luminaries and how we deal with those issues. The Average Rated Lamp Life, Initial Lamp Lumens, Mean Lamp Lumens, Lumen Maintenance at rated lamp Life, Ballast Factor, Ballast Efficiency, Thermal management, Optical System Losses and of Course Dirt Depreciation of the Optical System and the Degradation of the Materials that make up the optical system are the main issues of concern.
Normally these issues would be addressed in more of a training type session that would last several hours either here at Hadco or at a remote location and would at best accomplish only a very basic understanding of how all these work together. I have read many published articles by different LED manufacturers, LED Integrators and LED Luminaire Manufacturers. At best the LED manufacturers only understand the physics of how they make the LED and the integrators only understand how to power up and heat sink the LED but not how an LED should function and perform as a system so uniquely described as a LED Light Engine and or SSL Engine.
A) To make things a bit easier let’s start with eliminating some of these variable and base this discussion on 32watt - 175 watt Metal Halide:
1) Dirt Depreciation, assuming we have a clean optics system that is sealed from dust and moister
2) Materials Degradation, assuming we are working with a glass optical system that is not affected by the harmful 300-400nm wavelengths of Metal Halide
3) Ballast Efficiency, assuming the HID Ballast and the LED Driver System is at least 90%, remember that a standard 175 Watt Probe Start Metal Halide lamp/ballast system consumes 210 Watts on a standard constant wattage auto-transformer ballast, that calculates to be a ballast efficiency of only 83%, of course there are more efficient electronic ballast systems today which provide as much as 90%)
4) Ballast Factor and Driver Factor is “1″, assuming the HID ballast and the LED Driver can drive the lamp and or light engine to 100% of it’s rated watts
5) Initial Lamp Lumens, assuming we understand that in less than one year the Metal Halide Lamps have settled to 75% of their initial lumen output where as measurable losses of an LED System are in most cases insignificant.
6) Rated Lamp Life, assuming we understand that most Probe Start Metal Halide lamps at best have 2 years of reliable lamp life and that the newer Pulse Start Systems can provide up to 15000 hours (3.4 years at a 12 hr per day ontime) of lamp life. Remember Metal Halide rated lamp life is based on a mortality rate of 50%, or 5 out of 10 lamps still running at the rated lamp life.
B) Now that leaves us with the effects of Lamp Lumen Maintenance, Mean Lamp Lumens, Thermal Management and Optical System Losses:
1) Lamp Lumen Maintenance “varies” from one lamp manufacturer to the next and so does LEDs. You would really have to dig for this information since the real data is normally proprietary to the manufacturer, I have recently in my research discovered that Metal Halide lamp Lumen Maintenance is as low as 60% at rated life, once again depending on the lamp manufacturer and the ballast manufacturer. The Philips Lumileds are the LEDs that we prefer to use in our products are guaranteed a 70% LM to 50,000 hours, that’s about 11.4 years at 12 hours per day on time. The dilemma here is that how do you compare a 3.4 year lamp life at best for Metal Halide to the 11.4 year Lumen Maintenance cycle of LEDs. Remember, the LED does not have a rated life that the Lumen Maintenance is based on. The LED Lumen Maintenance is based on where the LED lumen output actually dims down to 70%. It is possible that LEDs can continue to operate for many more years rather than the Metal Halide 50% mortality rate at 3.4 years? We prefer to compare the LED Lumen Maintenance curve to the Metal Halide Lumen Maintenance curve which is roughly about 91% at 3.4 years which is where Metal Halide is rated. Given this comparison, if we were comparing an HID source and an LED source lumen to lumen, LED units would provide a 23% higher lumen output at 3.4 years than the Metal Halide Lamp.
Example: Bare 175 Watt Pulse Start MH Lamp/Ballast verses bare LED Iight Engine/Driver
a) GE Pulse Arc Lamp rated at 12,500 Mean Lumens with 90% efficient electronic Advance Ballast uses 194 watts at 120 Volts AC, that an efficacy of 12,500/194 = 64.43 lumens per watt
b) Cool White Lumiled Rebel Led rated 100 Lumens at 350 ma has an efficacy of 86.96 lumens per watt, lets be rally fair, we will deduct 5% for Junction Temperature losses then we have 82.61 LPW, and deduct
the 10% driver efficiency loss we have 74.35 LPW
c) In this simple comparison we have established that LEDs can provide 13.3% more lumen output bare lamp to bare light engine.
2) The next big concern is optical systems efficiency.
a) The typical HID lamp produces a toroidal luminous distribution out put pattern about the vertical axis of the lamp. Except for the typical 30 degree solid angle cone that is void of light at the ends of the lamp arctube luminous flux from the lamp radiates outwardly in all directions and not evenly.
b) The typical LED emits luminous flux from the silicon chip from the LED package outwardly in one direction into a solid angle of about 120 degrees. That makes LEDs “directional”.
c) The HID Flux distribution pattern would be an excellent source if a toroidal light distribution pattern was all we needed, but we know that this is not the case. The upset in the comparison between LEDs and HID light sources is this difference. The “real” gain in efficiency of LEDs over HID lamps is almost always in the efficiency of the optical system that gathers the luminous flux from the light source and redistributes into a useable lighting distribution pattern for particular applications whether it be a roadway scenario or a task oriented Luminaire such as a flood light. This is an area of the Luminaire design that can deliver up to 30% more luminous flux onto the work surface, especially if the area is a complex geometrical shape, liken to softball, football, soccer and basketball fields and also type 1, 2, 3, 4 and 5 IESNA lighting distribution. Try maintaining a sharp cut off with a single source Luminaire.
The real winner here in LED luminaries is “illumination” and not how big the lumen package is. I recommend to our sales reps that we’re selling illumination and not high lumen output lamps. A good example is gas mileage, I had a Dodge back in the 70’s with a V-8 and a 4-Barrel Carburetor that developed about 225 HP and delivered about 12 miles to the gallon on fuel efficiency. Today I have a Buick with a V-6 and is fuel injected that delivers about 225 HP at about 25 miles to the gallon. The real need today in an automobile is transportation and fuel economy. Today’s automobile manufacturers have figured out that “less” can be “more”. Just like with LEDs, the real story here is illumination and how we apply lumens.