Is Smart Lighting Energy Smart? Casper Kofod HD, M.Sc.Elec.Eng. Energy piano, L F Cortzensvej 3, DK – 2830 Virum, Denmark e-mail: [email protected]

Abstract Smart LED lamps appear more and at the market. These lamps provide the consumer with higher welfare, better product quality and/or energy and economy savings. Actually, most attention is paid to the first type of smartness which is the topic of this paper. The lamps are typically controlled by a smartphone app that enable the lamps can dim, change colour, shift to new scenes or programmed lighting changes over time. However, the lamps consume energy even when they are not in use. Often is also needed a separate energy consuming gateway for communication.to the lamps. The first indicative measurements from Australia, Europe and USA show the standby functions drastically increase the lamp’s total energy use and the standby energy can be larger than the energy used for providing lighting. In the future, there might be dozens of wirelessly controlled lamps in a single home, which taken together could result in high standby power losses. Therefor the IEA 4E SSL Annex has launched a new study on the energy performance of smart wireless lighting. Examination of the lamps as well as the communication protocols and gateways are expected to provide an evidence base for making recommendations to the government. The indicative measurements show design improvements for smart lamps are certainly possible in order also to become more energy smart.

Introduction Smart LED lamps have appeared at the market during the last years. The major part of these lamps provide better welfare by the ability to dim, change colour or make use of programmed scenes with gradual changes in the lighting. These smart lamps are wireless controlled by a smartphone app or a dedicated remote control. The capability of wireless control has the consequence that when the lamp is not emitting light, it switches to a standby mode waiting for a signal from the end-user to switch-on again. In a more advanced system the lamps may also be serving as part of a local wireless network, in a more active energy consuming mode. This means that these lamps are also consuming energy when they are not emitting light. Besides this standby consumption for the lamp and/or the luminaire there is often an extra consumption for a gateway which is a communication unit. These gateways - also called connection hubs or bridges - provide wireless communication between the general network for other devices and the lamps communication using protocols such as Zigbee, 6LoWPAN or Bluetooth. The IEA 4E SSL Annex [ref. 1] recognises that the market potential of smart lighting is extensive, once the platform enabling functionality has been established. Besides use in the home or at work, these smart lighting products can also be used in other applications such as museums, exhibition halls, shopping centres

and supermarkets where the lamps might be used as WiFi nodes to help consumers with smart phones to navigate a building or find products in a store. For example, the interaction between the smart lights and smart phones could activate visual and aural information for self-guided museum tours or detailed product information in a store. The opportunities and applications are virtually limitless. The risk that smart features may offset some of the energy-efficiency gains from switching to LED technology is the reason why the SSL Annex has started to study smart wireless lighting currently offered in the market, assess their energy use and the performance characteristics of these lamps and systems and evaluate possible appropriate measures to facilitate lower standby mode power consumption. There are other smart lamp features that prolong the lamp life (thermal control) or ensure to maintain a constant flux by regulation of the drive current. These features do typically not include standby consumption only extra control in the active state. Finally, there is different kind of sensors often of too low quality build into lamps or luminaires or placed external which typically are used for energy saving and improvement of the economy for the user. These two groups of other lamp smartness are outside the scope of this paper.

Smart Lamp Communication The smartness of the smart lamps is due to a user interface typically by use of a smart phone where the user can adjust the lighting e.g. the amount of light, the colour of the lighting, scene setting and programmed gradual change of the lighting. The communication from the smart phone or a dedicated unit is typically wireless provided by Wifi or Bluetooth or wired by Ethernet. The communication to the lamp often requires a lamp gateway

Figure 1 Smart lamp communication There is often need for a lamp gateway to convert the communication used to access the lamps and for communication between the lamps. The lamp gateway is thus something different from the general WiFi router used in a home. The lamp gateway (also called bridge or connection hub) might be:

1. 2. 3. 4.

A separate box with own separate mains power supply. Built into one lamp that can take care of communication to a group of lamps Built into every lamp Not necessary in case the user interface and the lamps are communicating by the same protocol e.g. WiFi

The protocol for communication between the lamps might e.g. be by Z-wave, Zigbee, 6LoWPAN, WiFi or Bluetooth.

Standby consumption The SSL Annex has compiled some test data on smart lamp products and tested in Annex Member Country lighting laboratories: • • • •

11 models purchased in the USA by Erik Page & Associates and tested at ITL Boulder, 3 models purchased in Europe and tested by the Swedish Energy Agency, and 2 models purchased and tested by the Australian government. 3 models purchased in Denmark and tested by DTU Photonic.

Some of the models in the different countries are the same. The sample sizes for these models tested are small (1-3 units) therefore the test results should at this stage only be interpreted as indicative measurements. Tests on this limited number of smart wireless LED lamps have revealed that the standby consumption for the different models varies between 0.17 and 2.71 W with most smart lamps having a standby consumption in the interval 0.4 - 0.6 W.

Figure 2 Standby consumption for smart lamps The lamp with a much larger standby consumption than the others includes a built in lamp gateway in every lamp. Some of the other models have separate lamps gateways. The average standby consumption for these separate gateways is 1.5 W. In the home, most lamps are used 1-2 hours/days [ref. 2]. The figure below shows the yearly consumption for the lamp models providing 225 – 1000 lm. When the operation time is 1 hour per day five of the lamps consume more than half of their annual energy consumption in standby mode. When the operation time is

increased to 2 hours per day only two of the lamps use half of the consumption in standby mode – anyway in this case the standby consumption is still considerable for 8 of the 9 lamp models.

Figure 3 Annual consumption for smart lamp providing 225 – 1000 lm The high standby consumption described above is similar to experience with standby consumption for other products where manufacturers first focused on the new features before turning their attention to reducing the standby power consumption. For example the load for smart phones chargers has for some models now been reduced to as little as 0.05 W.

Efficacy for the active mode and the total real total Efficacy The total energy consumption for the tested smart lamp models is much higher than for simple LED lamps of equivalent light output operated by the normal on/off switch. When the standby energy consumption is taken into consideration the efficacy is lower. The real efficacy for the smart lamps is calculated as shown below.

Below the efficacy only for the active mode is compared with the real efficacy.

Figure 4 Active mode efficacy and real efficacy for smart lamps providing 225 – 1000 lm

For the 9 lamp models, the active ‘ON’ efficacy varies within the interval 36 – 87 lm/W. With an operation time of 1 hour per day, the real efficacy varies within the interval 12 – 53 lm/W. For two lamps, the real efficacy is down at the level of incandescent lamps. In case of an operation time of 2 hours per day, the real efficacy varies within the interval 18 – 61 lm/W.

Decoration lamps Smart lamps with low lumen output also appear at the market. These lamps are mainly for decoration. They have low energy consumption but there could be many of them. For these lamps, the measurements show that that the standby consumption constitute a larger share of the total consumption than for the smart lamps with higher lumen output.

Figure 5 Annual consumption for smart lamp providing 50 – 75 lm For the decoration lamps, both the ON efficacy and the real efficacy are very poor. The real efficacy is less than half of the efficacy for incandescent lamps.

Figure 6 Active mode efficacy and real efficacy for smart lamps providing 50 – 75 lm

Need for metric for measurements and test of functionality The colour tunability has huge influence on the lumen output, the power consumption and thus the efficacy. Measurements at one of the most common smart lamp providing 600 lm, shows the efficacy is 13 lm/W in case the lamp is tuned to give blue light, 43 lm/W for red light, 110 lm/W for green light, 75 lm/W for warm white and 88 lm/W for cool white.

This complexity shows there is a need for standards for how to measure the performance for smart lamps. As it might take time before a standard is available for smart lamps, the SSL annex intend to develop a metric for smart lamp measurements as an interim methodology that can be used temporary. It seems also to a challenge to explain the consumers about the smartness of these lamps and the influence on their energy consumption.

The challenge for regulation The energy impact of smart lamps is difficult to predict as it is dependent on the functionality developed and the mode of deployment by the suppliers, as well as the user acceptance and usage of the functionality. Everything equal, the first measurements indicate smart lighting might increase the lighting energy consumption considerably as seen for other technologies and functions that involves network standby. Even with a small standby wattage, this might significantly increase the annual energy consumption. It seems thus desirable to regulate the standby consumption for smart lamps at an early stage to provide an envelope of consumption under which the functionality may be developed, but on the other hand this should not inhibit the potential for innovation. At present there appears to be no definition that adequately encapsulates the potential functionality, nor the network standby modes of these smart lamps, and consequently no suitable test methodology against which to regulate. This clearly presents a challenge for the regulator. A 'wait and see' strategy before attempting to develop suitable test methods and regulation is with high risk as the technology may be rapidly embedded as soon as the actual high prices start to decrease. Hence, development of an interim test methodology and associated regulation may be considered appropriate. For the voluntary US EnergyStar set of requirements has lately been discussed if it could be appropriate to require the standby power is below 0.5 W. In the ongoing revision of the EU lighting regulation which will come into effect within some years, the Danish Energy Agency has proposed to consider a requirement of maximum 0.3 W.

Summary Smart lamps combine technology breakthroughs in wireless communications and light emitting diodes (LEDs). The lamps are typically controlled by a smartphone app that enable that the lamps can dim, change colour, shift to new scenes or programmed lighting changes over time. All this are provided simply by touching your phone. However, the lamps consume energy even when they are not in use. The first indicative measurements show the standby functions drastically increase the lamp’s total energy use and the standby energy can be larger than the energy used for providing lighting. In order to better understand smart lighting features and their associated energy use, the IEA 4E SSL Annex has launched a new study on the energy performance of smart wireless lighting. Examination of the lamps as well as the communication protocols and gateways are expected to provide an evidence base for making recommendations to the governments. In the future, there might be dozens of wirelessly controlled lamps in a single home, which taken together could result in high standby power losses. Therefore, the IEA 4E SSL Annex hopes that their study already now will help to raise attention to the standby usage, so that design improvements can be made to the circuits leading to lower smart lighting standby power. Some lamps already operate with 0.17–0.25W standby power while most lamps use more than the double of this and in one case the standby consumption is ten times higher. So design improvements for smart lamps are certainly possible in order also to become energy smart.

References

[1]

The SSL Annex works internationally to support efforts at a national and regional level by addressing the main challenges with SSL technologies in order to develop a consensus on harmonised approaches to SSL performance and quality. The work of the SSL Annex spans a wide range of initiatives including guidance for policy makers, quality and performance tiers and support for laboratory accreditation, see http://ssl.iea-4e.org/. The SSL Annex is collaborating with the IEA EDNA (Electronic Devices and Networks) Annex which focuses on all kind of network connected devices including smart lamps, see: http://edna.iea-4e.org/about.

[2]

Assessment of the initial situation IEE/11/941/SI2.615944, Energy piano, 2013

in

the

participating

countries,

PremiumLight,