«

»

LEDs – Analysis and Results (Part V)

CHAPTER I – INTRODUCTION

CHAPTER II – LITERATURE REVIEW

CHAPTER III – DATA COLLECTION

CHAPTER IV – METHODOLOGY

CHAPTER V – ANALYSIS AND RESULTS

Analysis

Environmental Analysis

Humidity

Exposed bodies of water evaporate toward equilibrium, and aquariums are no different.  In humid environments (such as Florida), the excess humidity from an aquarium can cause mildew and rust problems.  In dry environments (such as Utah), the humidity can be a welcome addition in a home.  Regardless of the external environment, the evaporation from an aquarium can cause electrical problems in a lighting system if the airflow exchange rate is not sufficiently high.  (Bridges, 2013)

Salt Creep

Salt creep is what results when water splashed out of the aquarium evaporates and leaves behind a trail of salt.  Over time without good housekeeping, these salt deposits can grow quite extensive and infiltrate nearly every crack.  It will corrode electrical components, fasteners, and damage unprotected light bulbs.  Electrical and lighting components must be shielded, and all metals should be corrosion-resistant.  (Bridges, 2013)

Temperature

Most reef aquariums are kept around 74-80 degrees Fahrenheit; operation outside this temperature range can have devastating results, to include coral bleaching and fish respiration difficulty.  Metal halides are notorious for their heat output and usually require a chiller to counteract their output.  This heat generation can also decrease the life of surrounding equipment.  A lighting system should have a minimized heat output to keep the water temperature as stable as possible.  Most LEDs have a maximum efficient temperature rating of only 120 degrees Fahrenheit and require a heat sink with a fan to maintain this temperature.  (Bridges, 2013)

Intensity Analysis

Wattage per Surface Area

The amount of light that corals can receive is typically measured as photosynthetically available radiation and is measured in micromoles per square meter per second (μmol/m2/s).  Photosynthetically usable radiation (PUR) is a more valuable characteristic since it measures the intensity at the correct spectrum, but it is more expensive and difficult to measure, so it will not be included in this analysis.  Coral needs vary by the individual coral species (various species host various zooxanthallae species and pigments), the coral’s native location (Great Barrier Reef, Indonesia, et cetera), the coral’s native depth (where it settled initially before collection), water turbidity (high turbulence can scatter light while calm waters allow greater penetration), and even the coral’s health history.  Therefore, even though basic PAR requirements are known for most coral species and can be measured, there is no guarantee that the coral will respond as expected.  Additionally, PAR meters cost several hundred dollars for the entry-level models, which most hobbyists cannot afford.  (Bridges, 2013)

In order to bridge this knowledge gap, a few “rules of thumb” were developed or modified.  “Watts per gallon” was a standard entry-level guide for other lighting systems, and the recommendation was approximately three to five watts of lighting per gallon of the aquarium.  Since light intensity and spectrum decreases with depth, the “watts per gallon” recommendation was imperfect at best.  The “wattage per surface area” is no different.  It simply provides a correlation where lighting wattage may affect aquarium health.  Additionally, the starting parameter was determined by a survey with respondents from a wide range of experience levels.  Although the baseline “wattage per surface area” is a starting point, the author has verified that it is a sufficient baseline to successfully grow SPS.  (Bridges, 2013)

In order to determine a suitable baseline for “watts per surface area”, a survey was created and placed on several online saltwater aquarium hobbyist forums, including the Wasatch Marine Aquarium Society, Reef Central, and Nano-Reefs.  Each respondent was asked for their aquarium dimensions and their lighting wattage.  The average respondent (n=85) had a “watts per surface area” of 0.22 for an LED system (Figure 16).  This is much lower than for a typical metal halide system (0.4-1.0), which is reasonable, as LED systems are more efficient in their output for the wattage they consume.  (Bridges, 2013)

Figure 16:  Watts to Surface Area Survey Response

Figure 16: Watts to Surface Area Survey Response

Therefore, the “watts recommended” for an LED system can be determined:

SA = L * W

Where:

SA =    Surface Area

L    =    Length

W =    Width

Eq

Where:

WR = Watts Recommended

CF =    Conversion Factor (12 in2)

Wattage per Depth

As previously mentioned, coral PAR requirements vastly differ due to numerous variables.  However, certain species and groups of coral have similar saturation and photoinhibition points, which will allow wattage groupings.  Saturation is essentially level of light required for optimal photosynthesis.  Photoinhibition is beyond the saturation point, where light levels may be harmful.  (Riddle, 2007)

Large polyp Scleractinian coral tolerate around 110-350 μmol/m2/s, saturation and photoinhibition, respectively.  Small polyp Scleractinian coral vary widely, but they generally have a safe point between 275-400 μmol/m2/s, but some species’ photoinhibition point may be up to 700 μmol/m2/s or as low as 250 μmol/m2/s.  Soft corals have saturation and photoinhibition points around 200-400 μmol/m2/s, respectively. Clams appear to have no known saturation or photoinhibition points.  (Riddle, 2007)

With these light intensity variations in groups of coral, it is best to view an aquarium as a more compact version of the ocean.  Light intensity on the sand bed should be 100-150 μmol/m2/s, 150-300 μmol/m2/s halfway up, and 300-400 μmol/m2/s in the usable upper-half portion of the tank.  (Bridges, 2013)

Optics Analysis

Most three-watt LED chips are manufactured with a 120-degree spread (60-degrees to each side), which is sufficient for most shallow tanks (under 24 inches deep).  A tank from 24-30 inches may find that 80-degree optics will help prevent shadows, but they may cause a spotlighting effect.  Deeper tanks may use 60-degree or more acute angles to help focus the light. (Bridges, 2013)

Spectral Analysis

For the past few decades, the consensus amongst the reefkeeping community is that only blue and white bulbs are of a concern.  Although white bulbs contain the full visible spectrum, the ratio of blue to white light nearly negated the full spectrum effects of the white light.  The rationale was that red (and amber, yellow, green, et cetera) did not penetrate to coral collection depths in the ocean.  However, most coral in the hobby are collected by free-divers, who manually remove the coral from its base with a chisel and hammer.  Therefore, most of these coral were collected within 30 feet (9.14 meters) of water, which is within the full-spectrum wavelength penetration (Figure 17).  (Bridges, 2013; Karpenko, 2012)

Figure 17:  Wavelength Penetration to Water Depth Source: Adapted from Karpenko, 2012

Figure 17: Wavelength Penetration to Water Depth
Source: Adapted from Karpenko, 2012

In addition to the spectrum the coral needs to survive and thrive, other spectrums are necessary for the aesthetic appeal.  Coral fluorescence is important to many aquarists, and without the proper spectrum, the coral will not display its maximum coloration.  Coral pigments can absorb light and reflect the light back in a longer wavelength as a form of luminescence.  (Finet, 2005)

Infrared

Infrared wavelengths increase the temperature within the aquarium and are not known as part of the photosynthetically usable spectrum.  Since a stable environment (including temperature) is ideal, it is recommended that wavelengths above 700 nm be excluded where possible.  (Bridges, 2013)

Red Spectrum

The red portion of the visible light spectrum is one of the most debated wavelengths.  Chlorophyll a (Chl a) is a photosynthetic pigment that has a major role in zooxanthallae, and it peaks around 685 nm.  Corals also contain xanthophylls, a photoprotectant.  However, these xanthophylls convert blue-spectrum light into non-radiant heat.  Without a similar function for red-spectrum light, a coral may be overexposed and bleach. (Riddle, 2007)  Therefore, the lighting should contain a dimmable red light source in order to provide the required red wavelength but prevent overexposure.  

Yellow Spectrum

Although there are Yellow Fluorescent Proteins (YFP) that emit around 525-570 nm, they are quite uncommon and are not applicable.  (Riddle, 2009)

Green Spectrum

Discosoma Red (DsRed) is one of the five major coral pigments, and it focuses its excitation and emission around 561 and 620 nm, respectively.  (Riddle, 2009)  Green Fluorescent Proteins (GFP) are the most numerous and are excited around 500 nm and emit around 510-520 nm. (Riddle, 2009)     

Blue Spectrum

Cyan Fluorescent Proteins (CFP) excite around 450 nm and emit around 490 nm.  (Riddle, 2009)  This easily excited protein is a predominant reason behind the high actinic coloration of reef lighting, and it is likely a reason why full-spectrum LED systems are not in greater use. 

Violet (and Ultraviolet) Spectrum

Blue Fluorescent Proteins absorb in UV (around 380 nm) and emit around 448 nm.  (Cubitt, et al., 1999; Heim, et al., 1994)  Violet/Blue excitation of 400-450 nm can emit cyan to green fluorescence of approximately 490-509 nm.  (Gruber, et al., 2008)

Moon Phases

The moon has a 29.5-day cycle, is more red-shifted than sunlight, has an intensity under water of only approximately 0.5-1 μmol/m2/s under a full moon, and drops to no intensity under a new moon.  The moonlight spectrum is roughly composed of 55% red light, 30% green light, and 15% blue light.  (Riddle, 2012) To simulate this effect, a 30-day cycle is easier to program into the LED controller.  Cool blue light should ramp from 0-1% (off or on), green light from 0-2%, and red light from 0-4% over a 0-12 hour period each month.  (Bridges, 2013)        

Controllability Analysis

Dimming Function

LED dimming is controlled by two methods:  pulse-width modulation (PWM) and analog.  PWM provides more flexibility, but it is also more challenging to control and can cause an annoying flicker if not set to the appropriate frequency.  Analog control is easier, but it can cause extraneous heat.  (Bridges, 2013)   

Channel Control

LEDs can be controlled on one or multiple channels, depending on the purpose.  With multiple LED colors, each color can be individually controlled or grouped.  The LEDs can also be controlled by location, but this will increase the amount of controller programming required.  (Bridges, 2013) 

Interoperability Analysis

Tidal Simulation

With a rudimentary moonlight cycle and a small water capacity, the tidal simulation capability is limited.  Rather than programming a complex water movement cycle, variable rate power heads are sometimes preprogrammed.  For instance, the EcoTech MP40 power head features (in addition to other programs) a “Tidal Swell Mode”.  The flow starts in a left-to-right movement, calms down, switches to right-to-left flow, calms down, and then ends in a large surge (Figure 18).  (EcoTech, 2013)  The placement of these swells can be aligned with the moon cycle to reflect the higher tides around the full and new moon.

Figure 18:  EcoTech MP40 Tidal Swell Mode Source:  EcoTech Marine, 2013

Figure 18: EcoTech MP40 Tidal Swell Mode
Source: EcoTech Marine, 2013

Weather Simulation

Although storms can leave negative lasting effects on a reef, they do help clean the area of trapped waste and debris.  The same goes for an aquarium.  Simulated cloud cover can provide coral photosynthesis relief, reduce cooling required, and save energy.  LEDs and some drivers can produce simulated lightning effects (short bursts of high intensity light), but this is for no known purpose other than show.  Combining cloud cover (and lightning effects if desired) with the EcoTech’s “Nutrient Transport Mode” through programming the controller will mimic the beneficial effects of storms (Figure 19).  (EcoTech Marine, 2013)

Figure 19:  EcoTech MP40 Nutrient Transport Mode Source:  EcoTech Marine, 2013

Figure 19: EcoTech MP40 Nutrient Transport Mode
Source: EcoTech Marine, 2013

Sunrise and Sunset Effects

The simplest simulation for sunrise and sunset is to simply dim the intensity across the spectrum used throughout the day.  Aquarium lengths six feet and greater can simulate a true side-to-side sunrise and sunset effect, but this effect is lost on narrower tanks.  (Bridges, 2013) 

User Analysis

The aforementioned survey asked users why they chose LED lighting systems (Figure 20).  The lack of heat produced and the low power consumption were the two most-cited reasons (16%).  A long life expectancy (15%), low cost (14%), and color (11%) followed.  High efficiency (6%), controllability (5%), and dimmability (4%) were other leading reasons.  (Bridges, 2013)

Figure 20:  User Reasons for Choosing LED Lighting

Figure 20: User Reasons for Choosing LED Lighting

Of a more technical nature is the light intensity and spectrum over each user’s aquarium.  Up until now, there has been very little correlation data between invertebrate health and LED fixture characteristics.  Each user was asked for their tank dimensions and total lighting wattage.  The aquarium surface area (length x width) was then calculated, and the total wattage was divided by the surface area.  This is an imperfect characteristic as light intensity and spectrum quickly dissipate by depth.  However, it is a start for correlations.  The average user has 0.22 watts per square inch, but some users have as little as 0.05 and as high as 0.5 (Figure 21).  The user with the highest wattage per square inch may have invertebrates with high photosynthetic demand while the user with the lowest may have invertebrates with nearly no photosynthetic demand.  Without studying each aquarium’s demand, it is difficult to develop a better correlation.  However, the author has witnessed several healthy high-demand systems with approximately 0.22 watts per square inch, so this parameter is deemed feasible with some exceptions.  Additionally, if this parameter is too high, then the light fixture can be dimmed to a more suitable number.

Figure 21:  Relationship of Watts to Surface Area per Response

Figure 21:  Relationship of Watts to Surface Area per Response

Out of 85 respondents, 29 respondents stated they had an adverse coral response to LEDs (although several responses sounded as though they did not understand the meaning of “adverse.”)  Nine reported a bleaching event.  (A bleaching event is where a coral is stressed to the point where it releases its symbiotic photosynthetic algae (Figure 22).)  Interestingly, those respondents who claimed their corals bleached after switching to LEDs had the same watts to surface area number, 0.22, on average as other users (Figure 23).  This suggests another parameter may be at play, such as light spectrum.  In fact, 66% of those who experienced a bleaching event used a full-spectrum light system.

 

Figure 22:  A Bleached Coral (top) and Healing (bottom)

Figure 22:  A Bleached Coral (top) and Healing (bottom)

Increasing red light intensity and/or duration will cause corals to expel their symbiotic algae until they completely bleach.  Photopigments within most coral do not absorb red wavelengths and have not developed protection against this exposure since approximately 40% of red wavelengths are attenuated within the first meter of water (Riddle, 2004).  Therefore, it is possible that the users who reported bleaching may have exposed their coral to excess red light.  However, red light does have a role in the captive reef environment as moonlight is composed of more red than blue wavelengths and may help signal vertebrate and invertebrate spawning (Riddle, 2012).

Figure 23:  Light Intensity at Surface for Bleaching Responses

Figure 23:  Light Intensity at Surface for Bleaching Responses

Requirements Analysis

Goal A:  Cost

Objectives

A.1      The fixture shall reduce the electrical consumption from the lighting system by 50% from metal halide.  This is traceable to user analysis.

A.2      The fixture shall have a reduced maintenance schedule of every five years compared to yearly of other systems.  This is traceable to environmental and user analysis.

A.3      The fixture shall eliminate the requirement for active heat extraction (demonstrated by maintaining a constant set temperature, ±2 degree Fahrenheit) other than fan use.  This is traceable to user analysis.

A.4      The fixture shall have a cost breakeven point of at most three years.  This is traceable to user analysis.

Goal B:  Healthy Photosynthesis

Objectives

B.1      The fixture shall provide adequate PAR for coral (to include various species of Montipora, Acropora, Favia, Acanthastrea, Scolymia, Acanthophyllia, Echinophyllia, et cetera), which is defined as a minimum of 100 PAR on the sandbed (approximately 30 inches below the water surface.)  This is traceable to intensity analysis.

B.2      The fixture shall provide an adequate full-spectrum that closely replicates a various coral species’ needs.  With a lack of evidence, the sun’s irradiance at various depths of water may be substituted.  At a minimum, the light must be able to be varied for spectrum at 32.8 feet (10 meters), 65.6 feet (20 meters), and 98.4 feet (30 meters) for coastal waters (slightly turbulent).  This is traceable to spectrum analysis.

B.3      The fixture shall not include any ultraviolet (UV) or infrared (IR) LEDs.  These are defined for this project as less than 400 nanometers (UV) and greater than 700 nanometers (IR) peak.  Detectable UV and IR shall be minimized.  UV wavelengths shorter than 380 nm shall be shielded.  This is traceable to spectrum analysis.

B.4      The fixture and individual LEDs shall be arranged to minimize the amount of shadows.  This is traceable to intensity analysis.

Goal C:  Control Parameters

Objectives

C.1      The fixture shall be dimmable, which means the user must be able to change the fixture’s overall intensity, and the intensity of each individual color LED set, from 0-100% in 5% increments.  This is traceable to intensity analysis, user analysis, spectrum analysis, and controllability analysis.

C.2      The fixture shall have a vast user-selectable color spectrum, through the use of dimmable LED combinations based on Objective C.1.  This is traceable to user, spectrum, and controllability analysis.

C.3      The user shall be able to control the time (on/off) of the fixture locally or wirelessly through a distant computer, smartphone, and/or other electronically compatible device with an internet connection.  This is traceable to user analysis.

C.4      The fixture shall be able to simulate the spectrum and phase (intensity) of the moon for a given location.  This is traceable to spectrum, user, intensity, and controllability analysis.

C.5      The fixture shall be able to simulate a sunrise and sunset effect.  This is traceable to user and interoperability analysis.

C.6      The fixture shall be able to simulate a cloudy day and increase water turbulence to mimic a storm.  This is traceable to user and controllability analysis.

Preliminary System Design

The author’s main display aquarium has a surface area of 1152 in2, which requires at least 288 watts of LEDs, and results in a watts per surface area of 0.25 watts/in2.  This is slightly higher than the average watts per square inch in Figure 14, but the system is dimmable.

In addition to wattage, the color LED ratio will determine the spectrum.  For every 14 LEDs, one red (660 nm) LED should be included.  Two cool blue LEDs (475 nm) should accompany every five royal blue LEDs (450-455 nm).  A cyan/turquoise LED (495 nm) should accompany every red LED.  Two royal blue LEDs should match every one neutral white (4500 Kelvin) LED.  A violet (430 nm) LED accompanies every four royal blue LEDs, and a violet (405 nm) accompanies every eight royal blue LEDs.

Lighting penetration is also of concern, so optics should be used for tank depths greater than 25 inches.  80-degree optics should be used for deeper tanks up to 30 inches deep.  Deeper tanks, such as the author’s 31 inch-deep tank should use 40-60 degree optics as well as higher wattage LEDs (such as 5W rather than the standard 3W).

Proper heat sink design is essential, as heat is one of the greatest enemies of electronics.  Unfortunately, most easily available heat sinks do not provide specifications of thermal impedance or other characteristics.  Instead, they provide a number of LEDs that the size heat sink can supposedly handle.  This was used to determine a safety factor (SF), in addition to any that the heat sink manufacturer built in.  Fans were not accounted for in the analysis, so the SF will only be improved with the use of fans.

Considered Designs

Two 18-inch Fixtures

The first design considered was two 6×18-inch heat sinks with 296 watts of LEDs (Figure 24).  The LEDs used were in the following quantities:

  • 24 x Royal Blue (450-455 nm), 5W each
  • 16 x Neutral White (4500 K), 5W each
  • 8 x Violet (420 nm), 3W each
  • 8 x Cyan/Turquoise (495 nm), 3W each
  • 8 x Red (660 nm), 3W each
  • 8 x Cool Blue (475 nm), 3W each

Figure 24:  Two 6"x18" LED Fixtures

Figure 24:  Two 6″x18″ LED Fixtures

Depending on the LED source and heat sink design, the cost for this fixture design ranged from $693 to $713 (Table 16).  The higher-performance heat sink led to the higher cost, but its SF was only 1.01.  The lower-performance heat sink could not handle this design since its SF was 0.44.  Therefore, the $713 design was the only viable 18-inch option.

Table 16:  Bill of Materials for the Two 6"x18" Fixtures

Table 16:  Bill of Materials for the Two 6″x18″ Fixtures

 Two 20-inch Fixtures

The second design considered was two 6×20-inch heat sinks with 300 watts of LEDs (Figure 25).  The LEDs used were in the following quantities:

  • 28 x Royal Blue (450-455 nm), 5W each
  • 14 x Neutral White (4500 K), 5W each
  • 10 x Violet (420 nm), 3W each
  • 6 x Cyan/Turquoise (495 nm), 3W each
  • 6 x Red (660 nm), 3W each
  • 8 x Cool Blue (475 nm), 3W each

Figure 25:  Two 6"x20" LED Fixtures

Figure 25:  Two 6″x20″ LED Fixtures

Depending on the LED source and heat sink design, the cost for this fixture design ranged from $651 to $753 (Table 17).  A high-end heat sink option was not available, so the SF was only 0.48.  Therefore, this design was not feasible.

Table 17:  Bill of Materials for Two 6"x20" Fixtures

Table 17:  Bill of Materials for Two 6″x20″ Fixtures

Two 24-inch Fixtures

The third design considered was two 6×24-inch heat sinks with 300 watts of LEDs (Figure 26) in the same configuration as the second design in order to increase the SF.  The LEDs used were in the following quantities:

  • 28 x Royal Blue (450-455 nm), 5W each
  • 14 x Neutral White (4500 K), 5W each
  • 10 x Violet (420nm), 3W each
  • 6 x Cyan/Turquoise (495 nm), 3W each
  • 6 x Red (660 nm), 3W each
  • 8 x Cool Blue (475 nm), 3W each

Figure 26:  Two 6"x24" LED Fixtures

Figure 26:  Two 6″x24″ LED Fixtures

Depending on the LED source and heat sink design, the cost for this fixture design ranged from $760 to $764 (Table 18).  The higher-performance heat sink led to the slightly higher cost, and its safety factor was greatly improved to 1.33.  The lower-performance heat sink could not handle this design since its SF was 0.58.  Therefore, the $764 design was the only viable 24-inch option.  However, a dual 24-inch fixture would cause interference with other equipment in the aquarium’s canopy, so this design was avoided to prevent further reconfigurations.

Table 18:  Bill of Materials for 6"x24" LED Fixture

Table 18:  Bill of Materials for 6″x24″ LED Fixture

Four 12-inch Fixtures

The fourth design considered was four 6×12-inch heat sinks with 312 watts of LEDs (Figure 27).  The LEDs used were in the following quantities:

  • 32 x Royal Blue (450-455 nm), 5W each
  • 16 x Neutral White (4500 K), 5W each
  • 8 x Violet (420 nm), 3W each
  • 8 x Cool Blue (475 nm), 3W each
  • 4 x Red (660 nm), 3W each
  • 4 x Cyan/Turquoise (495 nm), 3W each

Figure 27:  Four 6"x12" LED Fixtures

Figure 27:  Four 6″x12″ LED Fixtures

Depending on the LED source and heat sink design, the cost for this fixture design ranged from $737 to $795 (Table 19).  The higher-performance heat sink led to the higher cost, but its SF was 1.28.  The lower-performance heat sink could not handle this design since its SF was 0.55.  Therefore, the $795 design was the only viable 12-inch option.

Table 19:  Bill of Materials for Four 6"x12" LED Fixtures

Table 19:  Bill of Materials for Four 6″x12″ LED Fixtures

With 60-degree optics on this fixture design, the aquarium should have sufficient light coverage with a peak intensity in the center of the tank (Figure 28 and Figure 29).  To reduce the peak and increase the lighting along the perimeter of the tank, the fixtures could be spread further apart.  However, this would increase the light spread outside of the tank, which is wasteful.

Figure 28:  Light Coverage with 60 Degree Optics

Figure 28:  Light Coverage with 60 Degree Optics

Figure 29:  Light Coverage with 60 Degree Optics, Side View

Figure 29:  Light Coverage with 60 Degree Optics, Side View

Detail Design and Development

Detailed Interface Design

The LED lighting system interfaces with the Neptune Apex controller.  As previously mentioned, the Apex uses proprietary software, but it is intuitive.  Basic programs were developed and are located in APPENDIX F:  Neptune Apex Code.  This programming has a specified time when the individual LED strings (Neutral White, Royal Blue, et cetera) turn on and off.  It also ramps the intensity throughout the day and night.  If the temperature sensors in the aquarium get too hot, the Apex will turn the lights off but leave the fans on.  If the aquarium gets too cold, it will turn the lights on as well as half the fans.  It can also control the “storm” program where the EcoTech MP-40 power heads increase the water turbulence and create clouds/lightning with the LED fixture.  It also simulates the sunrise/sunset feature of the LED system and can incorporate tidal patterns with the power heads.

Usability Testing

After the first build, the author tested each individual string of LEDs, combinations of LEDs, the intensity of the LED combinations, temperature variations, and controllability (see APPENDIX E:  Fixture Build Documentation).  The light fixture was initially dimmed over the aquarium at 15% and will be increased each week by 5% to meet the DOE.  This intensity-ramping will occur during Phase 3, as corals and other invertebrates can take up to six months to adjust to their new surroundings.

Usability Test Participants and Location

The author and William Bridges (author’s spouse) each individually tested the fixture, which is in operation on the author’s 150-gallon main display aquarium, located in South Weber, Utah.  Final test and evaluation will occur during Phase 3 over a six-month period.

Usability Results

The usability test and results will not be determined until Phase 3, which is outside the scope of this project.  This is due to the long time that invertebrates take to produce a noticeable response to a change in their surroundings.

Results

Requirements Results

The new fixture uses 72 LEDs, and each individual LED is capable of running up to 3-5 watts.  The fixture is currently consuming approximately 50 watts (not running at full capacity), including the fans, compared to almost 1000 watts of the previous metal halide fixture.  However, the lack of heat output from the LEDs has caused the aquarium heaters to run more often, which raised the overall energy use.  Figure 30 shows the current draw from both metal halides (on the left side) and LEDs (on the right side).  Even with the heaters, the LED fixtures have an overall energy consumption 41% less than the metal halides.  Therefore, the LED fixtures met Objective A.1.

Figure 30:  Current Draw from Metal Halides (Left) and LEDs (Right)

Figure 30:  Current Draw from Metal Halides (Left) and LEDs (Right)

Objective A.2 required that the fixture have a reduced maintenance schedule to once every five years.  LEDs have a life of 50,000 hours, and if they are run for eight hours each day, then their expected life is approximately 17 years.  This gives a life safety factor of 3.42.  The drivers are the highest risk, so additional fans were placed in the canopy to keep the drivers from overheating.  Theoretically, Objective A.2 was met, although further testing is required.

Figure 31 shows the temperature variation over seven days, with the metal halide variation on the left and LED variation on the right.  The sample standard deviation of the metal halide temperature was 0.53 degrees Fahrenheit while the sample standard deviation of the LEDs temperature was 0.15 degrees Fahrenheit.  The metal halide system standard deviation is typically much greater in the summer since the ambient air temperature is warmer.  Regardless, the LED system is influencing the aquariums temperature to within Objective A.3’s requirement.

Figure 31:  Temperature Variation with Metal Halides (Left) and LEDs (Right)

Figure 31:  Temperature Variation with Metal Halides (Left) and LEDs (Right)

Actual LED lighting system costs came to a total of $1132.40.  Table 20 shows a breakdown of the LED costs of $297.11, including spares.  Driver costs are shown in Table 21, and miscellaneous supply costs are shown in Table 22.  Driver costs were $206.00, and miscellaneous costs were $629.29.  The total parts cost was $1132.40, which was 5.6% below the budget of $1200.00.

Table 20:  LED Costs

Table 20:  LED Costs

Table 21:  Driver Costs

Table 21:  Driver Costs

Table 22:  Miscellaneous Supply Costs

Table 22:  Miscellaneous Supply Costs

The breakeven point for the actual first article LED fixture build was 29 months, which met Objective A.4.  However, the costs did not include labor, but they did include the initial purchase price, energy costs, and maintenance costs.  Figure 32 shows the breakeven points for various lighting systems and shows that the breakeven cost was $1675.  However, the breakeven point should be even sooner as Figure 32 assumes the light fixture is run at 100% capacity for a worst-case scenario.

Figure 32:  Breakeven Point for New LED System

Figure 32:  Breakeven Point for New LED System

Objective B.1 was met as the sand bed consistently measured 100-150 PAR with a Seneye Reef PAR meter.  The middle of the aquarium had PAR values ranging from 200-300 PAR, and the uppermost corals had PAR values of 300 to over 400.  Directly under the lights at the water surface, the PAR was 1300, which is comparable to high performance metal halide systems.

The spectrum analysis is not yet complete as tests are ongoing due to a faulty meter.  However, this does not impact Phase 3 testing.

Objective B.3 was met as no dedicated UV or IR LEDs were included in the build.  The 405 nm violet LED included does emit some UV radiation, but the amount is small (Figure 33).  Emissions below 380 nm are negligible.  Some IR radiation is emitted as well in the form of heat, but this was minimized by choosing the heat sinks with the highest safety factor.

Figure 33:  405 nm Violet LED Spectrum Source:  LEDGroupBuy.com

Figure 33:  405 nm Violet LED Spectrum
Source:  www.LEDGroupBuy.com

Figure 34 shows the finished LED fixture hanging above the author’s 150-gallon reef aquarium.  The individual heatsink components are hung in a slight parabolic shape in order to minimize shadows.  Some shadows could not be avoided due to the rock structure inside the aquarium.

Figure 34:  LED Fixture over 150-gallon Aquarium

Figure 34:  LED Fixture over 150-gallon Aquarium

Objective C.1 required that the fixture be dimmable from 0-100%.  At 0% intensity, the LEDs are off.  Unfortunately, with an analog controller and dimmers, the LEDs do not respond from 1-9%.  They turn on at 10% and are controllable from 10-100% in 1% increments.  This was a trade-off as lower-end control was sacrificed for higher-end control.  See APPENDIX F:  Neptune Apex Code for the code to control intensity.

Since all six colors of LEDs have intensity control from 10-100%, the number of color spectrum options available to the user is 5.3×1011.  This is more than sufficient to meet Objective C.2.  Standard color spectrums for user ease are still in development due to a faulty meter.  In other words, settings will eventually be available for the user to choose a 10,000-Kelvin, 15,000-Kelvin, or a 20,000-Kelvin spectrum with little effort.

Figure 35 shows the computer interface with the Neptune Apex, accessed remotely via wireless internet.  Figure 36 and Figure 37 show the cell phone interface on a Motorola smart phone.  The interface was also tested on an HTC EVO 3D smart phone.  From this dashboard, the user is able to see a snapshot of the temperature, pH, and current usage.  Additionally, the user can control all the LEDs at once or individual LEDs to adjust the spectrum.  The user also has control over the fans, heater, email alarms, and other parameters.

Figure 35:  Remote Neptune Apex Control Panel

Figure 35:  Remote Neptune Apex Control Panel

Figure 36:  Cell Phone Accessibility

Figure 36:  Cell Phone Accessibility

Figure 37:  Cell Phone Control Panel

Figure 37:  Cell Phone Control Panel

Objective C.4 (moonlight simulation) is still in development since low-end control was sacrificed for high-end performance.

Objective C.5 (sunrise and sunset simulation) was met as the fixture program ramps the colors and intensities up and down over an hour.  The code is in APPENDIX F:  Neptune Apex Code.

Objective C.6 was partially met as the fixture can simulate periods of high water flow with the Neptune Apex, but cloud cover is still in development due to the low-end intensity performance.

Test Results

Testing began on 9 November 2013 with the metal halide fixture and continued through 23 November 2013.  The first LED test started 24 November 2013 and continued through 7 December 2013 with an intensity of 15% Royal Blue, 15% Neutral White, and 15% Other Colors.  The subsequent weeks will consist of a ramping effect up to 15% Royal Blue, 15% Neutral White, and 30% Other Colors.  At that point, the testing will be outside the scope of ENM 590 and into Phase 3.

Specific gravity has stayed constant at 1.027 (Figure 38).  The vertical green lines represent changes in the lighting, and error bars on each data point represent the test’s accuracy.  There is no known correlation between salt uptake and coral growth, so a stable specific gravity is expected.

Figure 38:  Specific Gravity

Figure 38:  Specific Gravity

Alkalinity levels briefly dropped below target range (Figure 39), so the daily dosage was increased to compensate for the additional coral growth.  The vertical green lines represent changes in the lighting, and error bars on each data point represent the test’s accuracy.  Although the daily alkalinity requirement increased, it did not appear to increase in relation to lighting changes. The alkalinity supplement was dosed at 70 ml per day and is now dosed at 100 ml per day (Figure 40). 

Figure 39:  Alkalinity

Figure 39:  Alkalinity

Figure 40:  Daily Alkalinity Dosage Volume

Figure 40:  Daily Alkalinity Dosage Volume

The calcium level appeared to drop during one test with constant dosing (Figure 41), but a multitude of anomalies could have caused this.  For instance, a new batch of calcium may not have been fully stirred, the tester may not have fully stirred the reagent or mixture, et cetera.  The vertical green lines represent changes in the lighting, and error bars on each data point represent the test’s accuracy.  The daily dosing level at 225 ml per day is currently sufficient to maintain coral growth. 

Figure 41:  Calcium

Figure 41:  Calcium

Figure 42:  Daily Calcium Dosage Volume

Figure 42:  Daily Calcium Dosage Volume

Magnesium levels were maintained at 1360 ppm with 50 ml of supplementation per day until 7 December, where the levels reached down to 1200 ppm (Figure 43).  The vertical green lines represent changes in the lighting, and error bars on each data point represent the test’s accuracy.  To compensate for the additional uptake, daily dosing was increased to 270 ml.

Figure 43:  Magnesium

Figure 43:  Magnesium

Figure 44:  Daily Magnesium Dosage Volume

Figure 44:  Daily Magnesium Dosage Volume

Temperature, as previously discussed in Requirements Results, now has a standard deviation of 0.15 and stays within the target range (Figure 31).  pH is also within the target range (Figure 45).  Phosphate is not within the target range, which could affect coral coloration and growth negatively (Figure 46).  To correct this, additional skimming, water changes, and more frequent GFO changes will occur.  Free ammonia has stayed constant at 0.001 ppm, which is negligible due to the accuracy of the tester.  Ammonium varies throughout each day, but the average is around 13 ppm.  This amount is negligible due to the pH.

Figure 45:  pH Variation

Figure 45:  pH Variation

Figure 46:  Phosphate Levels

Figure 46:  Phosphate Levels

Although the LED light system is entering Phase 3 test and evaluation, the results are extremely promising.  At only 15% power, the LEDs are reaching equivalent metal halide performance at a much lower cost with more control.

Disclaimer

The views and opinions expressed or implied in this paper are those of the author and should not be construed as carrying the official sanction of the University of Dayton, the Engineering Department, or of individuals/groups mentioned in this paper.  This project is for informational purposes only, and it should not be used replace proper electrical engineering training before attempting such a project.  Any projects arising from this paper are at the reader’s own risk.  Additionally, this report and analysis shall not be used for commercial and/or profit without the author’s explicit written permission and any permission required from the University of Dayton.

References

Leave a Reply