INDUSTRY INSIGHTS | Evaluating energy impacts of 0–10V control in outdoor lighting

Although analog 0–10V control dates back to the 1980s when it was developed to dim fluorescent ballasts, it remains the most commonly available control option offered by North American manufacturers of LED luminaires and lighting controllers. Other control methods offer greater capabilities but can be proprietary, complex to configure, and costly to implement.

Significant tradeoffs can result from using 0–10V technology with the intelligence of LED and its modern network interfaces and luminaire-level sensors. Predicting relative luminaire light output and input power at any particular control voltage is difficult due to dependencies on LED driver design and loading, and in practice the performance across luminaires can be inconsistent.

The net result of unpredictable and varying luminaire responses to input control signals and a standard practice that does not compensate for, or even acknowledge, these variations is that end users regularly see unexpected and undesirable performance. Some examples include:

• Energy and cost savings estimates associated with illumination reductions may not be realized.
• Adjacent luminaires with different LED drivers may dim to different light levels when provided with the same control signal, which may compromise visibility or be visually unappealing.
• Dead zones in the dimming curve, where the control signal is changing but light levels are not, may lead users to think that one or more aspects of the lighting control system are not functioning as intended. 

Initial study of outdoor luminaires

Pacific Northwest National Laboratory (PNNL) has been studying this issue and reported on our preliminary findings in a prior article, “To dim or not to dim: Why is that still a difficult question?” The initial study examined the variations in response of commercially available streetlights to 0–10V control signals. Researchers looked at 23 LED outdoor cobrahead luminaires with ANSI C136.41 receptacles that claimed dimmability via a 0–10V interface and applied NEMA ANSI C137.1-2019 to address performance variations that may be found across the controls used to dim them.

Our evaluation made it clear that the relative luminaire power draw (as compared to rated) at a given 0–10V control signal voltage can vary significantly. The ANSI C137.1-2019 standard intended to reduce some of the variation seen in market-available products, and when we compared its requirements with the evaluated streetlights, a little more than half of them complied — indicating that specifications that require compliance with this voluntary standard should have an impact. However, testing showed that compliance with the standard only reduces average relative power draw variation from ~53 percentage points for all products to ~47 percentage points for compliant products. So, overall, while ANSI C137.1-2019 might have a positive impact on market-available products, it will not necessarily result in uniform or predictable performance.

The result of this lack of predictability can be significant. Outdoor lights that are dimmed in response to resident complaints or in order to execute control strategies may not deliver the expected lighting levels, thereby compromising resident satisfaction and possibly safety. Lights that are subject to adjustments in electricity costs based on expected reductions in energy use may be over- or under-billed. In city deployments with tens of thousands of streetlights, the variation in dimming performance could have a significant impact on expected energy use and cost.

A closer look at energy performance impact

The observed variation in response of streetlights to 0–10V control signals can have a significant impact on luminaire energy performance and cost if unique make and model products are not calibrated.

The research team quantified this impact by performing an energy analysis for an example medium-sized city installation of 20,000 streetlights. We looked at three lighting control strategies that modulate light output and power, and compared energy and cost savings against a baseline condition with lights operated all night at 100% of rated power. We assumed — as is commonly the case in the real world — that luminaire dimming curves were not measured in order to calibrate the 0–10V control signal response. In such cases, the most common expectation is that the relationship between control voltage and input power is perfectly linear between 0V and 10V, such that, for example, 10V and 5V control signals would deliver relative power levels of 100% and 50%. Actual dimming levels were drawn from the dimming curves measured from 23 streetlights described previously.

Two baseline conditions were considered — one in which the luminaire is operated “uncontrolled” and assumed to draw “full rated power” and a second in which the luminaire is fed a “10V control signal” and draws the power that we observed during characterization at that control voltage. Although these two baselines might be assumed to be equivalent, our characterizations show this not to be the case, as luminaire response to a 10V control signal varies.

The first lighting control strategy the research team considered was a simple part-night dimming (PND) strategy — where luminaires are dimmed to an assumed 50% of full rated power for 4 out of 11 hours of operation. We also looked at two constant light-output (CLO) strategies (L₉₀ at 20 years and L₈₀ at 16 years) that counteract the typical design practice of specifying luminaires that provide a minimum required light level at the end of their life, and consequently provide more than this minimum during normal operation, as they degrade from their initial output over some period in a predictable way.

For example, the lighting output of a luminaire with L₉₀ at 20 years’ degradation is expected to steadily decrease from 100% to 90% over the course of 20 years. CLO strategies initially dim luminaires to achieve the target light output, and then gradually increase output power to counteract the expected depreciation over time. In practice, such CLO strategies are the most commonly deployed control for outdoor lighting.

The PND strategy should deliver an annual energy and cost savings of 18%. However, based on the results of an analysis that compared “expected” savings with actual savings realized by each of the real-world measured dimming curves, the deployment of luminaires with 0–10V LED drivers will result in, on average, a savings of 12% or 13%, depending on the baseline condition.

The L₉₀ at 20 years CLO strategy should deliver a cumulative energy and cost savings of 5% at the end-of-strategy period. However, based on a similar analysis of expected versus actual performance, the deployment of luminaires with 0–10V LED drivers will result in, on average, a cumulative increase in energy use of 2% or a cumulative savings of 1% at the end of the strategy period, depending on the baseline condition.

The L₈₀ at 16 years CLO strategy should deliver a cumulative energy and cost savings of 10% at the end-of-strategy period. However, based on a similar analysis of expected versus actual performance, the deployment of luminaires with 0–10V LED drivers will result in, on average, a cumulative savings of 0% or 3% at the end of the strategy period, depending on the baseline condition.

The nearby figure shows cumulative energy and cost savings delivered by these three strategies for all dimming curves characterized in the study. Full details on these analyses and results can be found in a forthcoming report that is expected to be available on the DOE website this fall.

Recommendations for 0–10V

The results of our studies will hopefully help the lighting industry and standards-developing organizations better understand and possibly resolve the shortcomings of 0–10V products and consider what is best for the industry — namely, some means for delivering predictable and consistent dimming performance across all luminaires in a lighting system and thus guaranteeing the delivery of expected light levels and energy use. This could be achieved to a limited degree if, for example, driver manufacturers were to design products that comply with ANSI C137.1–2019. It could be achieved to a greater degree if the lighting industry further refined the standard to tighten up its requirements. Both driver and luminaire manufacturers could facilitate greater predictability by providing dimming curves for their 0–10V products as part of basic product documentation, and transition to standardized digital methods of control (such as DALI D4i or ANSI C137.4–2021) that do not have the same unpredictable performance and variation as 0–10V control.

Developers of connected lighting systems could enable the assignment of characterized 0–10V dimming curves to each deployed make/model luminaire in the system by their central management and lighting controller software, such that the 0–10V control signal sent to the luminaire can be “calibrated” to ensure that the luminaire draws the desired relative power and delivers the associated relative light level. In addition, developers could create mechanisms in their software for systematically capturing luminaire dimming curves by sweeping the control signal input and measuring the resultant luminaire input power draw — for luminaires or luminaire controllers that are capable of monitoring input power.

Standards development organizations could modify ANSI C137.1 to define a single 0–10V high control voltage, either 8V or 9V, and define a relative driver output power requirement or range at the single 0–10V high control voltage (for example, 100% of full output) and at a specified 0–10V low control voltage (such as 8%–12% of full output).

Lighting designers and specifiers could do their part by requesting dimming curves for all specified 0–10V luminaires and ensure that luminaires connected to the same control signal will perform consistently or within expectations. Designers and specifiers could also consider specifying DALI D4i and/or ANSI C137.4–2021 compliant drivers to achieve accurate and consistent dimming performance across all luminaires in the system and thus guarantee the delivery of expected energy and cost savings.

Working together, we can achieve more accurate and consistent dimming performance across all luminaires in a lighting system. Notably, while our work focused on LED streetlights, similar results are to be expected for other luminaires, as the underlying phenomenon is not a function of lighting application.

A more comprehensive reporting of this work, including more test method details and data analysis, will soon be available on the DOE connected lighting systems website.

ANAY WAGHALE, SHAT PRATOOMRATANA, MICHAEL POPLAWSKI, and JASON TUENGE work at Pacific Northwest National Laboratory (PNNL), where their research efforts primarily support the U.S. Department of Energy Building Technologies Office.