A New Horizon in Lighting: PFO's Solaris LED System

Major advancement has been missing in the aquarium lighting arena until PFO Lighting recently introduced their light-emitting diode (LED) system. The Solaris is one of the first widely available 'advanced' lighting arrays. The Solaris is available in several models and all offer some advantages over other lighting. I was fortunate enough to receive a beta model from PFO for testing. Although this prototype lacks some of the progressive features of the production models, it enabled me to evaluate a high intensity LED array. The prototype I received has 25 3-watt LED lamps (13 blue and 12 white) with an advertised Kelvin rating of 20,000 (see Figure 1).

Figure 1. The LED array of a 14" Solaris luminaire.

PFO makes some bold claims in their advertising. How well do these 'facts' hold up under close scrutiny? Several testing protocols were used to evaluate lamp intensity, lamp spectral quality and heat transfer.

Methods

Photosynthetically active radiation (PAR) was measured with a LiCor LI-189 quantum meter with cosine-corrected submersible sensor. This instrument is within its recommended 2-year calibration period. Calibrations were made for 'air' or 'water' as appropriate. An Apogee quantum meter (using the 'sunlight' mode) was also used to take measurements within an aquarium, and these results were compared to those of the Li-Cor meter.

A hand-held Project Star spectrometer demonstrated a visual display of light energy between 400 and 700nm, and was recorded by an Olympus C-5050 digital camera in 'macro' mode.

Spectral quality was analyzed with an Ocean Optics USB-2000 spectrometer and cosine-corrected CC-3 collection lens. Raw data from the spec was interpolated to 1nm increments and smoothed with the Savitsky-Golay algorithm. This information was further analyzed for color percentage in a proprietary Excel program.

Temperatures were measured with a 'laboratory grade' mercury thermometer and a laser-sighted infrared temperature 'gun'.

Light Intensity - Photosynthetically Active Radiation (PAR)

Light intensity, along with spectral quality and other factors, is an important part of captive coral maintenance. If the light intensity is too low, the zooxanthellaes' compensation point will not be met, sometimes with dire consequences. On the other hand, too much light will cause photoinhibition and zooxanthellae and the coral host can suffer.

PAR (photosynthetically active radiation) was measured 50mm below the protective lens situated immediately below the luminaire's LED array. Forty-eight measurements were taken and the results are graphed in Figure 2.

Figure 2. PAR measurements taken 50mm below the LED array.

Light Output Maintenance

An important selling point of LEDs is invariably that of lamp longevity. Philips maintains their white and blue Emitters have an expected life of 50,000 hours. The lamps are reportedly at 70% of original output at that point. If the daily photoperiod is 12 hours, then a 30% reduction in intensity could be expected in about 11 years. Figure 3 reports projected lamp intensity after 11.4 years of use.

Figure 3. Estimated light output after 50,000 hours operation (11.4 years at 12 hour per day photoperiod).

Solaris Intensity vs. that of a 250-watt 20,000K Metal Halide Lamp

PFO advertises the Solaris to produce as much PAR as a 250-watt 20,000K metal halide lamp. This is a difficult claim to verify - measuring and comparing the output of different light sources is, at best, a challenging proposition. Many variables come into play including type, age, production run and temperature of the light source, shielding material, lamp orientation, type of ballast, line voltage, lamp-to-sensor distance, type of reflector (including material, geometry, condition, etc.), type of sensor, effect of heat on the sensor and so on. I took time to control and maintain as many variables as possible, including lamp-distance-from-sensor, sensor temperature, etc. The XM 20,000K lamp was shielded for UV with an acrylic screen (the Solaris luminaire also has a 'splash guard).

Since these LEDs channel a relatively high proportion of input energy to visible light production, how does their efficiency compare to metal halides? We are comparing a metal halide lamp (a point source) to an LED array (multi-point source), and in order to fairly evaluate light production we should examine the light intensity over a broad but standardized area. Without going into a lot of detail, we can simply compare lamp wattage to PAR production over a given area. Using the data shown in Figures 2 and 4, plus standardized surface area, the 75-watt Solaris produced 89.4% of the PAR generated by the 250-watt XM 20,000K lamp.

Figure 4. PAR values of the 20,000K lamp (with acrylic UV shielding) used in one of the comparisons. The rather unique light distribution pattern is due to the reflector geometry.

This equates to totalized PAR (µmol·sec) over a normalized surface area of 113 µmol·sec per watt for the Solaris, and 38 µmol·m²·sec per watt for the metal halide lamp. See comments in the Discussion section. However, we should also consider Photosynthetically Usable Radiation (PUR).

Spectral Quality - Photosynthetically Usable Radiation (PUR)

Light intensity is only part of the picture, and PUR should also be considered. PUR is that fraction of PAR that is absorbed by zooxanthellae photopigments thereby stimulating photosynthesis. For our purposes, we will consider PUR as those wavelengths falling between 400-550nm (absorption bandwidth of chlorophylls a, c², and peridinin) and ~620-700nm (red absorption bandwidth of chlorophylls a and c²).

When all LEDs are on, their emission peaks at 458nm. Without the 4 LEDs that can act as 'moonlight' the spectral peaks shifts ever so slightly to ~459nm. This is well within the absorption bands of and other photopigments founds within zooxanthellae. See Figure 5.

Figure 5. The spectral signature of the Solaris when all 25 LEDs are in operation. Intensity is reported by the Ocean Optics software as 'counts.' The amplitude of the counts in Figure 5 and Figure 7 mean nothing - see the text for PAR measurements.

Figure 6. Visualization of the Solaris LED spectral quality. A hand-held spectrometer clearly shows that the combination of blue and white LEDs do indeed produce full-spectrum light. See Table 1 and Figure 5 for further refinement of this information.

Table 1. Spectral Power Distribution of the Solaris, expressed as percent.

Figure 7. A graphical representation of the SPD of the Solaris LED array.

The output of the XM 20,000K lamp peaks at ~450nm. The XM lamp's output is also efficient in producing PUR. See Figures 8 and 9 as well as Table 2.

Figure 8. The SPD of the XM metal halide lamp.

Table 2. Spectral Power Distribution of the 250-watt XM 20,000K metal halide lamp expressed as percent.

Figure 9. A graphical representation of the SPD of the 250-watt XM 20,000K metal halide lamp.

When we compare the spectral qualities of the LED array and the metal halide lamp, we see that the Solaris produces more PUR. See Figure 10.

Figure 10. Photosynthetically Usable Radiation of LEDs and metal halide lamp.

PAR within an Aquarium

This particular portion of testing presented more challenges than all others combined. The Solaris used in testing is but 12" long, making it suitable for small and nano-reefs. The LiCor PAR sensor is used as the gold-standard in testing, but the size of the sensor makes it of limited use in small tanks. The Apogee PAR meter with its smaller sensor makes it practical for use in smaller tanks - but its sensor is less accurate. Comparisons between the two meters found the Apogee's response is closer to that of the LiCor when the toggle switch is in 'sunlight' mode. Still it reads about 10% low when compared to the much more expensive LiCor meter. Correction factors were determined and applied to the results produced by the Apogee. The highest corrected measurement in 9" of water and directly below the middle of the Solaris luminaire was ~330 µmol·m²·sec. The Solaris does a good job of producing a tight beam (note the rectangular light pattern in Figure 2). PAR values fall off dramatically in the corners of the aquarium (data not shown). Placement of corals within the aquarium should therefore not be haphazard but done instead with forethought and planning. Branching corals (Acropora spp., among others) and Tridacna clams should be placed directly below the luminaire, while shade-loving animals such as Discosoma and others will probably do well outside the direct light beam.

Heat Transfer

Heat transfer from a lamp to an aquarium's water can cause significant warming.

PFO advertises that all heat generated by the LED array is directed away from the aquarium. To test this claim, the Solaris was placed atop an aquarium and about 2" from the water's surface. The Solaris was allowed to operate continuously for about 27 hours, and temperatures were measured by an infrared heat gun and a 'laboratory' thermometer. In all cases the water temperature did not exceed room temperature (maintained at 24.3ºC by room air conditioning). The top of the luminaire is vented, and it did warm to 25.5ºC while the bottom (lamp side) stayed at room temperature. It appears that PFO's claims are accurate.

Discussion

PFO selected and uses one of the best LEDs on the market - the Philips Lighting 3-watt Luxeon Emitter LEDs. Selection of LEDs for use in an aquarium fixture is critical and, again, PFO has done its homework. The 'batwing' design of the Philips Emitter series allows for superior thermal management around the diode since it allows the junctions to act as thermal sinks and permits airflow for cooling. Heat is an enemy of LEDs since it can shorten their expected lifetime. The Solaris luminaire utilizes two cooling fans to maintain air flow across the circuitry. PFO claims the electronics within the luminaire are encapsulated with a material impervious to the inevitable salt spray. Unprotected circuitry was an issue with some of the LED banks produced 5 or so years ago. PFO has apparently overcome this problem and their protective circuit coating should go a long way in preventing corrosion.

The LEDs are mounted in Fraen 'wide beam' lenses and housings. The dispersion angle is 42º for the blue LEDs and 45º for the white ones. At first, the use of the wide beam lenses seemed puzzling until I took some PAR measurements. It appears as if use of the narrow or medium lenses would be too efficient and create hotspots of high-intensity light. The wide beam seems correct for most, if not all, home aquaria applications.

Lighting requirements of many, if not most, corals is overestimated by many hobbyists and lower light intensity offer advantages. This is why. Many corals are stressed by high light intensity and have natural means to deal with excessive radiation, either through dynamic or chronic photoinhibition. Chronic Photoinhibition, from the coral/zooxanthellae perspective, is by far the worse of the two - it could eventually kill zooxanthellae. Obviously, we want to avoid this. On the other hand, dynamic photoinhibition involves biochemical conversion of excess light energy to non-radiant heat. This is normal and does no apparent harm to either coral or symbiont. However, it is a waste of light and the electricity needed to power the lamps. And you'll pay for it every month. In addition, few hobbyists would want to operate a hot (>260°C, or >500°F) metal halide lamp just two inches from the water's surface. The halide lamp will definitely transfer heat to the aquarium. Unless properly shielded, metal halide lamps can produce high amounts of detrimental ultraviolet energy. The Solaris produces practically no UV radiation and - for our purposes - it is safe to say it produces none. Blue LEDs have been shown to trigger expression of a pink coloration (a 'pocilloporan') in the stony coral Pocillopora meandrina.

Very low heat production is perhaps the greatest advantage of LEDs. Circulation pumps will add more heat than this lighting system, but this can be overcome easily through use of a relatively large sump or by airlifts (which are practical in systems compatible with the small size of the 14" Solaris luminaires). Aquaria large enough to accommodate larger Solaris models will also benefit from lessened lamp-to-water heat transfer.