LED lighting and control systems evolve for optimal efficacy (MAGAZINE)

New SSL system designs and control architectures will allow LED-based lighting to better realize the efficacy potential of the technology, explains JOE DENICHOLAS.

Content Dam Leds En Articles 2011 07 Led Lighting And Control Systems Evolve For Optimal Efficacy Magazine Leftcolumn Article Thumbnailimage File
+++++

This article was published in the July/August 2011 issue of LEDs Magazine.

View the Table of Contents and download the PDF file of the complete July/August 2011 issue.

+++++

For multiple fiscal and environmental reasons, lighting efficacy – defined loosely as light only when, where, and how it is needed – should be given the utmost consideration when we deploy lighting systems. From an energy-consumption standpoint, LED-based lighting represents the most important advancement in lighting in decades. LEDs as light sources are inherently efficient and LEDs can be configured in systems that are much more intelligent in terms of both controllability and adaptability than traditional fluorescent and HID technologies. Indeed LED-based solid-state lighting (SSL) can provide an advantage in efficacy from many angles, but luminaire and control-system architectures must evolve to deliver truly optimal efficacy.

From a system perspective, lighting efficacy is comprised of several elements, all of which are of first-order importance. Several are outlined in Table 1. Light source efficacy is not enough. Truly efficient lighting also requires efficient electronics, fixtures that don’t waste light, and control systems that further reduce wasted light.

Efficacy element Description Units
Source efficacy Ability to convert electricity to visible light. May or may not take into account photopic, mesopic, and/or scotopic human visual system response, and even Color Rendering Index (CRI) Lumens per watt (lm/W)
Power supply efficiency Power source to load conversion Percent (%)
Fixture efficacy / light distribution efficacy Light directed to target relative to wasted light, and distribution uniformity as portions of the target area may need to be over-lit to achieve minimum required levels elsewhere in the pattern Lumens per watt (lm/W) as calculated according to Fitted Target Efficiency
Utilization factor/ supply vs. need Over-lit conditions, due to lack of occupancy sensors or user preference waste energy with little to no incremental user benefit and sometimes user detriment Percent (%)

Table 1. Elements of lighting efficacy

Based on efficacy advantages, LED-based fixtures appear to be either in the lead or quickly approaching the lead in many applications such as high-bay lighting, street lighting, indoor downlighting and even fluorescent troffer replacement. Still, we need to rethink proper light levels, focus on lighting only where it is required, and push deployment of control schemes to maximize energy savings and eliminate light pollution in the environment.

The lighting industry still has work to do in determining proper light levels. For example, regulatory bodies in North America do not currently take into consideration the differences between photopic (day), mesopic (dusk), and scotopic (night) human visual systems. Our visual system has evolved to account for the differences in lighting between day and night. During bright sunlit days, our eyes are more excited by warmer CCTs (correlated color temperatures) than during dim nights when our eye sensitivity shifts toward the colder, more-bluish moonlight. Mesopic lumen output describes a situation in between photopic and scotopic and is generally considered the most appropriate measure for street lighting.

Efficacy and eye sensitivity

The differences in efficacy can be dramatic when considered relative to photopic, mesopic, and scotopic sensitivity. This is shown in Table 2, which compares a low-CCT high-pressure-sodium (HPS) source to a much-higher -CCT, metal-halide (MH) source. High-CCT sources such as MH and LED are not necessarily given proper credit for exciting the eye in an optimal way for given environmental conditions. Given the data in the table, it’s no surprise that many people involved with case studies report that LED street lights with a lower total lumen output appear brighter than higher-total-lumen HPS street lights. Note that this statement refers to the brightness of the target area and not the fixture itself which may (falsely) appear brighter due to glare effects. We need standards that ensure safety without wasting light and energy.

SourcePhotopic efficacy (lm/W) Mesopic efficacy (lm/W)Scotopic efficacy (lm/W)
HPS (low CCT)12597 78
MH (high CCT)107155175

Table 2. Comparison of high-pressure sodium (HPS) and metal-halide (MH) source efficacies.

Likewise, some regulations and guidelines don’t consider the CRI (color rendering index) of a light source even though it has recently been proven to have an effect in some applications (again, like street lighting) where small-target visibility is critical. Both CCT and CRI are critical because the required lumen output of a lamp varies greatly based on these factors. That said, their importance is still being debated and as recently as 2007, CIE’s stated position in CIE 180:2007 is that, “Colour rendering is not highly important for roadway lighting, except in sensitive urban centres and/or areas with large numbers of pedestrians.”

Utilization factor

Now let’s discuss utilization factor. The first three efficacy elements in Table 1 are static, at least within a relatively short timeframe of days or weeks. This is not the case with the fourth element that addresses the difference between the light supplied relative to the light needed. Utilization factor is a combination of the percentage of time that the lights are on and, when lights are on, the intensity of the light compared to what’s required or being utilized. Optimized lighting controls are essential to improving utilization factor and thereby reducing energy costs. LED lights present a new opportunity for controls as they are easy to regulate using various dimming methods, sensor interfaces, and communication infrastructures that allow the light to be modulated based on environmental conditions.

Lighting systems can perform occupancy detection to control on and off states. Several technologies can detect occupancy including passive infrared (PIR) or ultrasonic motion sensors, capacitive- or MEMS-based microphones, and digital cameras that perform image processing. Motion sensors are relatively inexpensive and are used most often although a combination of a motion sensor and another occupancy-detection method can yield superior performance. Multi-technology sensors decrease the likelihood of erroneous behavior, thus maximizing precision and decreasing energy usage.

Controlling fixtures and dimming lights to produce the appropriate amount of artificial light based on ambient light conditions is critical to both energy efficiency and user experience. Dual-loop sensors are now able to differentiate between light provided by the sun and artificial lighting systems so that fixtures can maintain a consistent light level on a target area. LED-based lamps have the advantage that deep dimming is easy to do and actually increases lamp life, in contrast with competing technologies.

Leveraging lumen depreciation

SSL also affords the potential of further energy savings in luminaire designs that accurately account for lumen depreciation in regulating light output. Light-output regulation is very important to LED-based lighting because of the technology’s extremely long lifetime. If properly protected and driven, LEDs shouldn’t burn out. Instead, the LED light output decreases over time based on a phenomenon called lumen depreciation. L70 is a parameter that describes the point in time at which the light output has decreased 30% from its initial value, and is typically on the order of 35,000 to 100,000 hours for LED lamps, as shown in Fig. 1.

FIG. 1.
To maintain a minimum amount of light output over the lifetime of a fixture, say 750 lm for a 65W replacement lamp or 6,000 lm for a parking-lot light, many fixture designs initially output 30% more light than is required. This represents a significant waste of electricity in that the target area is being over-lit for virtually the entire lifetime of the fixture.

Intelligent fixtures can regulate the light output to a lower level initially and increase the output over the fixture life. Ancillary benefits include consistency of light intensity and color, lower overall energy expenditure, and lower total thermal load. Lowering the total thermal load is extremely beneficial as it leads to longer lifetimes for all electronic components, especially the LEDs and power electronics.

Though beneficial, light-output regulation provides a significant technical challenge. One could use a predictive algorithm that estimates LED efficacy or output based on hours of operation and temperature measurements. But LED performance over time and temperature may not be all that predictable. For several families of LEDs from various suppliers, the actual lumen-depreciation curves have been shown to be significantly shallower than those predicated by accelerated, high-temperature testing.

Alternatively, a fixture design could add a sensor to measure the lumen output during operation, but there are challenges here as well. First, achieving proper mechanical placement of the sensor to measure overall- or average-lumen output may be difficult or even impossible. Second, dirt can can prevent photons from getting out of the fixture and may even redirect them towards the sensor, thus corrupting the measurement. Third, sensor aging and temperature drift could complicate matters even further.

FIG. 2.
In lighting systems, external sensors could measure the light output and communicate the data to the fixture. Such a system could be cumbersome, costly, and have its own set of technical issues. The right answer is likely a combination of approaches, and light-output regulation appears to be one area that is ripe for innovation.

Microcontrollers and networks

Clearly the industry must move toward intelligent lighting platforms to maximize energy savings via sensors, programmatic controls, and communications links between fixtures. Such intelligent luminaires rely on driver modules that integrate a microcontroller for interfacing to sensors and for control of the dimming profile. The smart fixtures enable managed-lighting systems with wired- or wireless-communications capabilities.

The communications infrastructure allows lights to communicate with each other, with remote sensors, and with centralized control and data-collection points. Such control systems have existed for some time but have not been widely deployed, having an estimated market share at 2% to 4%. Cost and complexity have hampered deployments. Moreover, the lighting industry focused first on more efficient sources such as fluorescent and HID that weren’t inherently controllable.

With LED sources, it’s time for broader deployment of control networks although the technology landscape is fragmented. Wired communications options include 0-10V dimming, DALI (Digital Addressable Lighting Interface), DMX (Digital Multiplex) or power-line communications. Wireless personal area network (PAN) options include Zigbee, Z-Wave, 6LoWPAN, or even Google’s new Android lighting platform. All may find usage although the market will likely pick the winners.

New lighting system topology

The trend is clearly toward systems that integrate the control strategies and intelligence directly into the ballast or driver. But, the overall power-supply and control architectures of these systems will likely change to take full advantage of LED technology. For example, consider a space lit by four 25W downlights, as shown in Fig. 2.

The lamps are controlled by remote occupancy and ambient-light sensors over a wireless PAN. A wired configuration could just as easily have been shown. Regardless, each fixture operates from line voltage and includes significant intelligence and therefore requires:

  • 25W AC/DC converter
  • 25W DC/LED constant-current converter
  • Radio for the wireless PAN
  • A relatively expensive microcontroller including flash memory for the PAN protocol stack
  • Energy meter
  • Optional sensors (temperature, light output, or color).

FIG. 3.
As shown in Fig. 2, data gathered by the MCU could be backhauled to a central location that records energy usage. Such a system could also be under remote control in addition to being able to interface to local sensors. This system, while perfectly functional, is expensive to implement and does not take into account the simple but significant fact that we now have a light source that is easy to power remotely. An alternative approach is shown in Fig. 3.

In this case, the 100W power supply incorporates the room controller/coordinator and is therefore capable of communicating directly with the sensors and the remote-control/data-backhaul interface. In this case, each fixture contains:

  • 25W DC/LED constant-current converter
  • A relatively inexpensive microcontroller
  • Optional sensors (temperature, light output, or color).

From a power-supply standpoint, one 100W AC/DC converter is both more electrically efficient and less expensive than four 25W AC/DC converters. Energy metering is performed at the centralized power supply instead of at each lamp. The lamps communicate with the 100W supply over an extremely simple and inexpensive wired interface and therefore contain a less-expensive microcontroller, lighter communications-protocol stack, and no radio. Finally, if the optional local sensors aren’t needed, then no electronics are required locally inside the lamp – the 100W power supply could send a constant current directly to the lamp.

Our proposed system lies somewhere in the spectrum between 100% local power supplies and intelligence and 100% remote power supplies and intelligence (something akin to Redwood Systems’ technology). The market must decide on the best solution.

Finally, artificial-intelligence or fuzzy-logic technology will enable these systems to become more efficient by enabling active learning – prediction of occupancy and even a user’s desired light level. Such systems could also greatly simplify and possibly even eliminate the commissioning process. This is obviously yet another area begging for innovative solutions.

More in Retail & Hospitality