LED Industry Overview

8/3/2019 LED Industry Overview

1/48

Sally Lee, LC

Applications Marketing Manager(& self-proclaimed lighting enthusiast)

OSRAM Sylvania, Inc.

LED Illumination inArchitecture

NTAEENovember 19, 2009Luncheon Meeting


2/48

| LED Color in Architecture | Background | Life and Efficacy | Applications | Specification Information |

Why is LED Lighting Important?

According to the DOE, energy efficient LED lighting is a solution to our energycrisis:

U.S. DOE has chosen energy efficient LED lighting to play the key role in reducing ourelectric light consumption by 50% by 2025

Over the next 20 years, rapid adoption of LED lighting in the U.S. can:

Reduce electricity demands from lighting by 62%

Eliminate 300 million metric tons of carbon emissions/year

Avoid building 133 new power plants

Anticipate financial savings that could exceed $200 billion/year

Slide: Lighting Science Group


3/48


LED illumination in architectureNext Generation Lighting Initiative (NGLI) -

EPAct 2005, Title IX

Secretary of Energy (DOE) to carry out a Next Generation Lighting Initiative (NGLI) tosupport research, development, demonstration, and commercial application activities forSSL.

$50 million to the NGLI for each fiscal year 2007 through 2009

authorization to allocate $50 million for each of the fiscal years 2010 to 2013

Public R&D investment serves the ultimate goals

to successfully commercialize the technologies in the buildings sector (lighting = 39%of total electricity use)

realize the full promise of solid-state lighting by 2025

speed SSL technologies to market unique attributes of SSL technologies underscore the importance of a long-term,coordinated approach encompassing applied research and strategic

technology

commercialization support


4/48


LED Illumination in ArchitectureSSL technologies are undergoing rapid change and improvement.

New SSL products are being introduced to the market almost on a daily basis.

Understand the lighting application considerations

Placement

Performance

Light level, luminance, contrast, glare

Products arriving on the market exhibit a wide range of performance and quality, particularly in the areas o

Color quality and consistency

Luminous efficacy

Lifetime

Value

Application efficacy

Total cost of ownership

Environmental impact

Emotion

Product comparisons are laborious: It is a huge task to compare LED options 1:1

Source: http://www.netl.doe.gov/ssl/PDFs/SSLTesting-OpenPre-QualFinal.pdf


5/48


What is LED (Light Emitting Diode)?

A solid object that emits visible light when energized

Semiconductor chips convert energy directly into light

Produces light without filament or arc

Compact Size (Fixture Design Flexibility)

No Infrared or UV emissions

Directional Light Source

Robust (No Moving Parts or Glass)

Low Voltage

Instant On and Dimmable

Wide range of color options

L igh t Emi t ted Forw ard

Tj

depends on drive condition and

application environment

Higher Tj

= Shorter life

(generally

life for every 10C)


6/48


White LEDs

Color Issues

Courtesy: Lumileds

Source: U.S. Department of Energy


7/48


0

50

100

150

200

2004 2006 2008 2010 2012 2014 2016 2018 2020Year

Efficacy(lm/W

)

DOE Published LED Component Efficacy RoadmapInstant-on, 350mA (March 2008, from Figure 4-7)

Commercial Product

Projection CCT > 4500K

Commercial ProductProjection-CCT 2700K-4500K

$20/ klm

$10/ klm

$5/ klm

$2/ klm

(Cost of LED component to generate 1000 lm)

HID: ~100FL: ~80-100CFL: ~40-80HAL: ~15INC: ~6-12


8/48


Facts in LED: Components vs. Systems

Factors 100% Efficacy 75-80% Efficacy

Color Temperature >4500K 2700K-4500K

Color Rendering low CRI < 80 high

CRI> 80

Fact 1

LED Component Efficacy Characteristics

Fact 2

LED System Efficacy Characteristics

Optics70%-90% Driver Electronics70%-90%Heat sink80-90% System Efficiency40%-75%

LED Lamp 400 lm (40W A19) LED Component 533lm

1000lm

Cool White, CRI < 80 LED Component Cost $11 -

$20

Warm White, CRI >80 LED Component Cost $13 -

$25


9/48


LM-79 (IESNA Guide for Electricaland Photometric Measurement of

Solid-State Lighting Products)

Tests the luminaire as a wholeasopposed to traditional testingmethods that separate lamp ratingsand fixture efficiency.

There are two main reasons for this:1) there is no industry standard test

procedure for rating the luminous fluxof LED devices or arrays

2) because LED performance is

temperature sensitive, luminairedesign has a material impact on theperformance of LEDs used in theluminaire.

System Efficacy = Lamp Lumens X BF / System Input Wattage

Luminaire Efficacy = Lamp Lumens Emitted / System Input Wattage

Facts in LED: Components vs. Systems


10/48


Application Efficiency

more useful light for applications ideally suited to work with optics


11/48


Guide DOE planning for SSL R&D and support

commercialization activities, including ENERGY STAR

program planning

Support DOE technology procurement activities and

associated technology demonstrations

Provide objective product performance information

to the public in the early years, helping buyers andspecifiers have confidence that new SSL productswill perform as claimed

Guide the development, refinement, and adoption of

credible, standardized test procedures andmeasurements for SSL products

The department will allow its test results to be distributedin the public interest for noncommercial, educationalpurposes only. Detailed test reports will only be providedto users who provide their name, affiliation, andconfirmation of agreement to abide by DOE's NoCommercial Use Policy.

Source: http://www.netl.doe.gov/ssl/PDFs/SSLTesting-OpenPre-QualFinal.pdf

Overall, fewer than 25% of the products testedhave met manufacturers

claims.

DOE CALiPER: Commercially Available LED Product Evaluation and Reporting

System Efficacy Improving Over Time

http://www.netl.doe.gov/ssl/comm_testing.htm
http://www.netl.doe.gov/ssl/comm_testing.htmhttp://www.netl.doe.gov/ssl/comm_testing.htm


12/48



DOE CALiPER:

Commercially Available LED Product Evaluation and Reporting

System Efficacy Improving Over Time


13/48


D.O.E. Technology Snapshot July 2009


14/48


More LED Activism: Lighting FactsLuminaire efficiency is not a well-understood concept by consumers


15/48


More D.O.E. LED Activism: The L Prize Competition

NOTE: the requirements prescribe specific end-user purchase price targets for these products over a 3-year period


16/48


17

DOE workshops: Began in March 2006,

many workshops since

Completed:

Photometric Measurement (IES LM-79)

Lumen Depreciation (IES LM-80)

Chromaticity (ANSI C78.377)

Definitions/Nomenclature (IES RP-16a)

Numerous other standards in various stages

of development

High priority for DOE

Standards and Test Procedures Support

www.ssl.energy.gov/standards.html
http://www.ul.com/http://www.ansi.org/


17/48



18/48


White LED Color Issues

General illumination:

CCT: 2800 K to 5500 K

CIE (x,y) values closeto the blackbody locus

Good color rendering properties

LED Lighting Institute, Feb 8-9, 2006


19/48


Influences on Color

Wavelength of die (460-470nm)

Amount/type of phosphor

LED Manufacturing Influences on Color

Blue wafer mapped forWavelength distribution


20/48


White LED

Color Issues

Color VariationSimilar light sources could have perceivable colordifferences between them.

Large color variation between similar lightsources is an undesirable feature.



21/48


Decorative / Architectural

applications require high color

homogeneity

if...

... several light points

can be seen directly

The human eye also detects color

differences easier at low brightness

... small beam angles

are used

... there are small distances

between object and lightsource

wheneverbeams dont get mixed

... A white surface

is illuminated

the whiter

the more sensitivee.g.

wall washing, cove lighting

or

a cluster of spots

White Binning for Color and Intensity


22/48


White LED

Color Issues

What is the MacAdam Ellipse?

In the early 1940s Dr. David MacAdam

tested the threshold sensitivity of the eye to

chromaticity differences in the matching of colored lights. MacAdam

found that the

threshold could be described with ellipses on the chromaticity diagram.

For the average human eye, color changes within a 4-step MacAdam

ellipse are

undetectable.

Dimensions of MacAdam ellipses in SDCM

(Standard Deviation of Color Matching)

Size of MacAdamEllipse

1 SDCM /

MacAdam Step

2-3 SDCM /

MacAdam Step

>4 SDCM /

MacAdamStep

Quality of color

homogeneity withinthe ellipse

No color

differencevisible

Hardly any

color differencevisible

Color

differencevisible


23/48


White LED

Color issues

How much color variation can there be?

1931 CIE Chromaticity Diagram

MacAdam EllipsesRepresent the loci of just-noticeablecolor difference

ANSI SpecificationCalls for 4-step MacAdam ellipse for

certain types of lamps

Wyszecki and Stiles, Color Science,

1982


| LED C l i A hit t | B k d | Lif d Effi | A li ti | S ifi ti I f ti |


24/48


White LED

Color Issues

Color Variation Between Similar 4100K Fluorescent

COMPETITIVE COLOR SPECIFICATIONS - ANSI 4100K

0.368

0.370

0.372

0.374

0.376

0.378

0.380

0.382

0.384

0.386

0.388

0.390

0.392

0 .3 70 0 .3 72 0 .3 74 0 .3 76 0 .3 78 0 .3 80 0 .3 82 0 .3 84 0 .3 86 0 .3 88 0 .3 90

"ANSI 4100K (.380,.380) - 4035K

2-step

4-step

OSI T8 (.382,.384) - 4010K

OSI T12 (.380,.378) - 4021K

GE SP41/SPX41 (.376,.387) - 4191K

PHILIPS SPEC/ULTRA (.376,.374) - 4103K

OSI DULUX (.380,.377) - 4013K



25/48


White LED

Color Issues

Color Variation Between Similar LEDs

White LEDs of similar type have noticeable colordifferences between them.

Color binning is one option to minimize variation.

What are the binning criteria?

15-step MacAdam Ellipse




26/48


White LED

Color Issues

New white LED binning

0,28

0,29

0,30

0,31

0,32

0,33

0,34

0,35

0,36

0,37

0,38

0,39

0,40

0,41

0,42

0,43

0,29 0,30 0,31 0,32 0,33 0,34 0,35 0,36 0,37 0,38 0,39 0,40 0,41 0,42 0,43 0,44 0,45 0,46 0,47 0,48 0,49

6500

K5700

K5000

K4500

K4000

K3500

K3000

K2700

K Take ANSI bins as they are

ANSIbins

| LED Color in Architect re | Backgro nd | Life and Efficac | Applications | Specification Information |


27/48


C IE 1 93 1 x,y C h rom a t ic i ty D iag ram

0 .28

0 .30

0 .32

0 .34

0 .36

0 .38

0 .40

0 .42

0 .44

0 .46

0 .2 6 0 .2 8 0 .3 0 0 .3 2 0 .3 4 0 .3 6 0 .3 8 0 .4 0 0 .4 2 0 .4 4 0 .4 6 0 .4 8 0 .5 0 0 .5 2

x

y 7-s tepM a c A d a m e l lip s e s

( C F L )

Il lum inant A

D 65

P l anc k i an

l oc us5000

K

6000

K

7000K

250

0K

3000K

4000

K

In c a n d e s c e n t W h ite ( 2 7 0 0 K )

W a r m W h ite ( 3 0 0 0 K )

C o o l W h ite ( 4 0 0 0 K )

S u n W h ite ( 5 0 0 0 K )

D a y lig h t ( 6 5 0 0 K )

W h ite ( 3 5 0 0 K )

C oo l W h it e -2 (4500 K )

S u n W h it e -2 (5 7 0 0 K )

Energy Star SSL Chromaticity Requirements

ANSI C78 377A Standard



28/48


Example of a Binning Scheme

0.30

0.35

0.40

0.45

0.

28

0.

33

0.

38

0.

43

0.

48

3000K

3500K

4000K

4500K

5000K

5700K

6500K

CD

EF

GH

I J

KL

MN

OP

QR

S TU

6

7

8

5

4

9

V W X

4

5

6

7

8

ANSI C78.377quadrangles

4000K

3000K

2500K

Planck

3step McAdams ellipses



29/48


9U

8U

7U

6U

6U

4U

9V

8V

7V

6V

5V

4V

9W

8W

7W

6W

5W

4W

9X

8X

7X

6X

5X

4X

9R

8R

7R

6R

5R

4R

9S

8S

7S

6S

5S

4S

9T

8T

7T

6T

5T

4T

9O

8O

7O

6O

5O

4O

9P

8P

7P

6P

5P

4P

9Q

8Q

7Q

6Q

5Q

4Q

8L

7L

6L

5L

4L

8M

7M

6M

5M

4M

8N

7N

6N

5N

4N

8J

7J

6J

5J

4J

8K

7K

6K

5K

4K

8I

7I

6I

3000 K

3500 K

4000 K

4500 K

0.33

0.34

0.35

0.36

0.37

0.38

0.39

0.40

0.41

0.42

0.43

0.44

0.45

0.46

0.47

0.35 0.36 0.37 0.38 0.39 0.40 0.41 0.42 0.43 0.44 0.45 0.46 0.47 0.48 0.4

To add, remove or edit the plots press 'Strg' and click on the chart

White LED -

Color Issues



30/48



White LED

Color Issues



31/48


White LEDs

Color Issues

Color Rendering Index (CRI)CRI is notaccurate for LEDs

**especially RGBs

New color quality metric underdevelopment

Long term effort

In the meantime use CRI as one data

point

Make decisions about color using thelamps that will be used in the finalinstallation




32/48


White LEDs

Color Issues

Some cracks in CRI discussion

- 60

- 40- 20

0

20

40

6080

100

R1R1 R2R2 R3R3 R4R4 R5R5 R6R6 R7R7 R8R8 R9R9 R10R10 R11R11 R12R12 R13R13 R14R14

Test Colours

There have been tests showing that people canprefer the light from low CRI LEDs than muchhigher CRI fluorescents and incandescents,

especially with saturated (a.k.a. bright) colors

0.0000

0.0050

0.0100

0.0150

0.0200

0.0250

350 400 450 500 550 600 650 700 750 800

Wavelength (nm)

TotalS

ectralFlu

x

W

nm

-1)

Control

Daylight

CRI Test Colors Special Test Colors

http://www.amerlux.com/pages/121_supermarket.cfm


33/48


Amplitude ratios

= 615, 525, 465 nm

CCT (K) CRI Red Green Blue

3200 72 .58 .25 .17

4000 73 .50 .26 .24

CRI variation as a function of peak wavelength shift at 2800K

LED wavelengths (R-615nm, G -525nm, B-465nm)

0

10

20

30

40

50

60

70

80

90

100

420 440 460 480 500 520 540 560 580 600 620 640 660

Peak Wavelength (nm)

CRI

CRI - Red Shift CRI - Green Shift CRI - Blue Shift

Small wavelength shift results in large CRIchange

Amplitude ratios

= 640, 525, 465 nm

CCT (K) CRI Red Green Blue

3200 21 .68 .21 .11

4000 27 .62 .21 .17

White LEDs

Color Issues

CRI Estimates RGB LEDs




34/48


White LEDs

Color Issues

Color Mixing Approach

White light source

White can be obtained by mixingtwo monochromatic sources

Undesirable color properties

Blackbody locus

Luminous efficacy and color properties are interrelated

PP

Wavelength




35/48


Measuring Light Source Life

The lifetimes of conventional light sources are rated through established test procedures.

Compact fluorescent, for example, is published by the Illuminating Engineering Society (IES)as LM-65.

Tests a statistically valid sample of lamps

ambient temperature of 25 degrees Celsius

operating cycle of 3 hours ON and 20 minutes OFF

The point at which half the lamps in the sample have failed is the rated average life

Business Question: For 10,000 hour CFL lamps, what is

total elapsed time (cumulative, in months)

that this process takes?

Conventional Sources: Time Unit of Failure



36/48



The lifetimes of LED SYSTEMS depends on

Application

Heat (thermal design)

drive conditions (electrical design)

useful light.

**Switching frequency is not an issue with SSL

Driving the LED at higher than rated current will increase relative light output but decrease

useful life.

Operating the LED at higher than design temperature will also decrease useful lifesignificantly

Business Question: For 50,000 hour LED lamps, what is total elapsed time(cumulative, in years) that this process takes?

LED: Beyond conventional Time Unit of Failure



37/48

| | g | y | pp | p |

Measuring Light Source LifeThe useful lifetime

of LED SYSTEMS

Useful illumination (L50

& L70

): IESNA L70

Thermal Design

Electrical design

Driving the LED at higher than rated current willincrease relative light output but decrease usefullife

Application (ambient temperature)

Operating the LED higher than designtemperature will decrease useful life significantly

Bonus: Dimming and switching frequency are notissues with SSL life



38/48

| | g | y | pp | p |

Defining LED Useful LifeAt what point is the light level no longer meeting the needs of the application?

May differ depending on the application:

Research has shown for general lighting in an office environment, that the majority of

occupants will accept:

light level reductions of up to 30% with little notice (L70

)

particularly if the reduction is gradual

Based on this research, the Alliance for Solid State Illumination Systems andTechnologies (ASSIST), a group led by the Lighting Research Center (LRC),recommends defining useful life as the point at which light output has declined to

1.

70% of initial lumens (abbreviated as L70) for general lighting

2.

50% (L50) for LEDs used for decorative purposes.

3.

They also note that for some applications, a level higher than 70% may be required.

http://www.netl.doe.gov/ssl/usingLeds/general_illumination_life_defining.htm



39/48

| | g | y | pp | p |

Measuring Light Source Life: ONE MORE THING

Useful life procedure-6000 hours of data

SYSTEM should FIRST be operated for 1,000hours, then monitered

while operating at rated

current and voltage for at least an additional5,000 hours in a ventilated environment at 25C.

For systems intended for operation where heatbuild-up would occur, the LED system should betested in environments corresponding to theirapplication ratings.

When color shift is critical, it is proposed thatchromaticity coordinates NOT fall outside of aMacAdam

four step ellipse (L4M) of initial.

SEEK THE SYSTEMS

LIFE VALUE

L70% (hours): Time to 70% lumen maintenanceL50% (hours): Time to 50% lumen maintenance

Within these times the LED component or system

should not exhibit chromaticity shifts greater thanthose bounded by a four-step

MacAdam

ellipse



40/48


What is an LED driver?An LED driver

performs a function

similar to a ballast for discharge lamps.

It controls the current flowing

through the LED.

Most LED drivers are designed toprovide current to a specific device orarray.

Since LED packages and arrays arenot presently standardized, it is veryimportant that a driver is selected thatis matched to the specific device or

array to be illuminated.

THE LED driver is a key componentto System performance and life.

http://../My%20Documents/Desktop/term238http://../My%20Documents/Desktop/term238http://../My%20Documents/Desktop/term238http://../My%20Documents/Desktop/term238http://../My%20Documents/Desktop/term238


41/48


Why is it important to control the currentthrough an LED?

A typical voltage-current relationship for anillumination-grade LED is shown in Figure 8.

As seen in this figure, a slight change in voltagecan result in very large changes in current.

Since the light output of an LED is proportionalto its current, this can result in unacceptablevariation in light output.

If the resulting current exceeds limitsrecommended by the manufacturer, the long-termperformance of the LED can be affected, resultingin shorter useful life.

Figure 8



42/48

CALiPER tests: luminaire temperature and expected life

There is currently no industry standard

which defines a case or board temperature

measurement point or other clear mechanism for determining junction temperature

for an LED when installed in a luminaire.

Thermal management is a key issue for solid state lighting.

The draft IESNA LM-80 (for

lumen depreciation measurement) currently describes a thermocouple attachment point

on LEDs, but it is unlikely that this measurement point would be

accessible once an LED

is installed in a luminaire.

CALiPER tests luminaires non-invasively primarily following the draft IESNA LM-79 andalso includes temperature measurements on hot spots

on the luminaires.

These external temperature measurements cannot be quantitatively

correlated to junctiontemperatures, but they provide some indication regarding the operating temperatures in

the luminaire.

CALiPER is also conducting lumen depreciation testing of selected luminaires--particularly

those whose thermal management may be insufficient or that may be subject to high 'insitu' operating temperatures (such as enclosed fixtures and recessed downlights).



43/48

LED Luminaire Performance:

Changing traditions to set the right expectations

Source: ledsmagazine.com January/February 2008 LED luminaire performance:

changing traditions can set the right expectations. (Narendran, Fryssinier, Taylor)



44/48

ENERGY STAR

Program Requirements for

Solid State Lighting LuminairesEligibility Criteria Version 1.0 (Final 9/12/07)

Recessed downlights -

Application Requirements

Minimum Light Output

4.5

Aperture: 345 lumens (initial)> 4.5

Aperture:

575 lumens (initial) Zonal Lumen Density RequirementLuminaire shall deliver a minimum of 75% of total lumens (initial) within the

0-60

zone (bilaterally symmetrical).

Minimum Luminaire Efficacy

35 lm/W

Allowable CCTs

2700 K, 3000 K and 3500 K for Residential products

No restrictions for Commercial

Reduced Air Leakage Recessed downlights intended for installation ininsulated ceilings shall be IC rated and be

leak tested per ASTM E-283 to demonstrate no more than 2.0 cubic feet per minute (cfm) at 75 Pascals

(1.57 lbs/ft2) pressure difference. The luminaire must include a

label certifying airtight

or similar designation t show air leakage less than 2.0 CFM at

75 Pascals

when tested in accordance with ASTM E283

.



45/48

ENERGY STAR

Program Requirements for

Solid State Lighting LuminairesEligibility Criteria Version 1.0 (Final 9/12/07) continued

Category B

is established as a future target for performance of SSL-based

luminaires. Products cannot qualify under Category B until three

years after the

effective date of these criteria. If technology improvesfaster

than expected, and

opening category B prior to completion of a three-year waiting period would be inthe public interest, DOE will advance the date for allowing products to qualify under

Category B.

Future (B) Luminaire Efficacy Target:Luminaire Efficacy

70 lm/W

All Other Requirements:Glare requirements to be developed

All other requirements will be the same as those in effect for Category A at the time

Category B becomes effective, except for minimum light output and zonal lumendensity requirements, which will not be used in Category B.



46/48

LED illumination in architectureLuminaire manufacturers have to deal withLEDs

sensitivity to thermal and electrical

conditions

Performance and efficiency claims madeby product vendors often greatly exceedknown start-of-the-art performance levels

Differences between LED sources and

conventional sources have created a GAPIndustry groups, standards organizations,and DOE are moving quickly to developneeded SSL standards and test procedures

In the meantime, there is a need forreliable, unbiased

product performance

information to foster the developing marketfor high-performance SSL products

Watch for color variations

Know the chip and driver entailed (IP clearance?)Does the system feature LEDs that have beenIESNA-LM80 tested (6000 hours; L70extrapolated)

Make luminaire comparisons using ABSOLUTE

photometeryIs it clear to you that the luminaire or lampmanufacturer participates in the DOE QualityAdvocate Pledge program?

What EXACTLY is covered by the warranty and by

whom?

For the lighting specifyer: The systemmanufacturer should provide a range ofpermissible operating temperatures within which

acceptable operation will be expected



47/48

Remember All the Pieces of the Puzzle

Lightingrequirements

Prototype

LED type

Optics

Thermal

management

LED driverPCB layout

Mech. design

Brightness


48/48