Headline figures for the PAR38 competition are a light output exceeding 1350 lm with a power of less than 11W, and an efficacy greater than 123 lm/W. Products should have a minimum center-beam luminous intensity of 18,000 cd. Not all lamps submitted for testing are required to meet these targets – see “PAR38 criteria” below.
The L Prize competition was established by US Congress in the Energy Independence and Security Act (EISA) of 2007 and launched by DOE in 2008.
The first L Prize was awarded to Philips in the 60W-incandescent-replacement category in August 2011. The winning Philips product was due to be available in stores in early 2012, and is in fact now being sold for $59.99 on Bulbs.com.
Philips received $10 million for winning the 60W-replacement competition. The EISA legislation authorized a $5 million prize for the PAR38 replacement, which the DOE says it has set aside from fiscal year 2012 appropriations.
The PAR38 replacement category was temporarily closed in 2011 to “retool the competition based on lessons learned through the 60W competition,” according to the DOE.
The relaunched PAR38 competition retains the original technical requirements established by Congress (see below), but has been “streamlined to keep pace with the speed of technology innovation and to move winning products into the market sooner,” says the DOE.
PAR38 lamps (PAR stands for parabolic aluminized reflector, and 38 designates a diameter of 120 mm or 4.75 in) are directional lamps that are in widespread use in both residential and commercial applications. The DOE says there are approximately 90 million PAR38 lamps installed in the US. Replacing these with lamps meeting the L-Prize criteria would save 11 terawatt-hours of electricity per year – approximately equivalent to the annual electricity consumption of Washington, DC – and avoid 7 million metric tons of carbon emissions.
“The L Prize competition challenges the best and brightest engineers and scientists across America’s lighting industry to drive innovation in new, more efficient products and boost our nation’s competitiveness in manufacturing,” said Energy Secretary Steven Chu. “The winning products will help expand lighting choices for consumers, reduce our nation’s energy use, and save money for American families and business owners.”
According to the DOE, “the rigorous performance testing needed to win the L Prize ensures that the performance, quality, lifetime, costs, and availability of winning products meet expectations for mass manufacturing and widespread adoption.”
One stated goal of the L Prize is to generate jobs for US workers. For the PAR38 category, at least 50% of the LEDs must be produced in the US, and all of the assembly must be done in the US.
PAR38 criteria
Full specifications and requirements for the PAR38 replacement competition are available from the L Prize website.
Among the changes made to the requirements are that only 1000 samples need to be submitted (versus 2000 for the 60W-replacement competition); companies must pay for their own LM-79 testing; and only 32 samples (versus 200) will be subjected to photometric and elevated-temperature lumen-maintenance testing.
For many of the product specs, only 90% of sample lamps need to meet the stated levels. For example, the initial light output is given as 1350 lm, but more accurately this means that at least 90% of units must exceed 1350 lm, and no unit is allowed to have less than 1215 lm.
Selected criteria:
- Light output: 1350 lm or more (10% of samples can be between 1215 lm and 1350 lm)
- Power consumption: 11W or less (10% of samples can be between 11W and 12.1W)
- Efficacy: greater than 123 lm/W
- CCT: 2750K or 3000K
- Lumen maintenance: 25,000 hours to L70 (70% of initial lumen output)
- CRI (Ra): at least 90 (10% of samples can be between 87 and 90)
- Dimmability: not required
- Center beam luminous intensity: at least 18,000 cd (10% of samples can be between 16,200 and 18,000 cd)
- Driver mean time to failure (MTTF): more than 25,000 hours at 45°C ambient temperature
- Power factor: at least 0.7 (10% of samples can be between 0.63 and 0.7)
