º Orchids Indoors
º Orchids Outdoors
º Greenhouses
º Culture Techniques
º Culture Specific
º General Articles
§ = member's only content
Additional Information
º Orchids Indoors
º Orchids Outdoors
º Greenhouses
º Culture Techniques
º Culture Specific
º General Articles
§ = member's only content
Additional Information

How to Choose an Aquarium Chiller
Choices for aquarium chillers are bewildering and available information is often missing important bits and pieces. Variables include lowest temperature that the unit will chill to (some go down to 32 F [0 C]), maximum operating temperature (our home can reach 90–95 F [32–35 C] inside during peak summer days), required flow rates, availability of external temperature probes to control the unit, energy consumption (in HP) and thermal power (in BTUs; 1 HP = 12,000 BTU) and noise. The common application is to cool a reef aquarium filled with water from more than 80 F to 78 F (27 to 26 C), and tank sizes for that application are in the more than 40 gallon (151 L) sizes. As my project is so different I enlisted the help of a specialized aquarium store to sort through the options. For instance, air has a much lower heat capacity than water. In terms of weight it is 1.005–1.84 kJ/kgK [0–100 percent relative humidity] for air vs. 4.187 kJ/kgK for water; with respect to volume it is on the order of 0.001–0.002 kJ/lK for air to 4.2 kJ/lK for water, hence a factor 2,000–4,000 difference. After checking out the parameter space on a webbased calculator (http://chiller.jbjlighting.com/prod_chiller_size.asp), they recommended a JBJ Arctica 1/15 HP chiller.
The approach taken was based on the expected heating of the water by the T5 lights. Although there is a total of 320 W of light applied to the overall aquarium, the water feature only occupies about 5 percent of its footprint. Accordingly, only 16 W of heat from the light is affecting the water temperature. Under the assumption of cooling the terrarium from 100 to 85 F (38 to 29 C), a 0.05 = 1/20 HP chiller was determined by the online calculator. On the other hand, the fan blowing over a relatively large surface area of water exchanges heat with the remainder of the terrarium air and structures. The latter factor could not be calculated beforehand, so it had to be approached experimentally.
Alternatively, one could consider the heat capacity of 70 percent RH air being approximately one third of that of water. Then we can use a third of the terrarium volume, but apply the full heat input from the light fixture (320 W). Including in the consideration the cooling of the light fixture by a fan, and assuming a cooling of the terrarium from 100 F to 85 F (38 to 29 C), a 0.07 = 1/14 HP chiller was recommended by the online calculator. With the compressor active about half of the time, the whole setup could cool the tank from 85 F to 76 F (29 to 24 C) with water cooled to 60 F (16 C), in about an hour. The 1/15 HP cooler is just about the right size—it neither cools the water to solid ice in a minute, but it manages to cool the water sufficiently for the present application. It appears that the cooling water does not need to be literally ice cold; in my particular case, a 15 F (9 C) temperature differential is sufficient, but this is likely to depend on the specifics of the terrarium and environmental conditions. There is a temperature gradient along the 5foot (1.5m) terrarium of about 2 to 4 F (1 to 2 C). To limit the absolute maximum temperature in the terrarium, place the temperature probe as far away from the swamp cooler as possible, or make allowances for this temperature gradient within the tank.
The conclusion from the practical experiment is that the online calculator seems to overestimate the chiller power possibly by a factor of two. On the other hand, chiller prices increase modestly with increased power, so a slightly oversized unit may be better in the long run than a slightly undersized one that needs to be replaced.
Light Levels for Plants
Outdoor light intensity is weaker by a factor of four to five in the early morning and late afternoon, compared with the peak hours (http://naturalfrequency.com/articles%5Caveragehourly), and varies by a factor of two to three over the course of the year (http://www.solarpanelsplus.com/solarinsolationlevels/) at intermediate latitudes. In a terrarium, lighting is mostly provided by a constant, artificial light source, usually on a 12hourson, 12hoursoff cycle at full output.
Plot of actual light intensity values. Note that the shape of intensity curve is in between a parabola and a Gaussian curve. (Data from www.naturalfrequency.com)
To approximate the appropriate light intensity for a constant terrarium light source, we can convert the total amount of natural light received over the course of a day to that provided by a simple on–off artificial light source. We convert the area under a curve to the area in a rectangle with same base length. The two areas represent the total amount of the light over the period of a day. The peak height of the curve is the peak daily illumination under natural light, whereas the height of the rectangle is the intensity of constant terrarium light source. By dividing one over the other, we obtain the ratio of peak intensity to constant illumination level. We can use three model curves:
º First, we can assume the light intensity increasing and decreasing along the outline of a semicircle. The area of a semicircle is πr2/2. A rectangle with length 2r and unknown height = light intensity h represents the sudden on–off switch with constant illumination levels in between, and has an area of 2rh. We equate the two areas πr2/2 = 2rh and solve for h = πr2/4r = πr/4 = 0.79r. The light intensity from a constant light source should be about 79 percent (~ 4/5) of the peak intensity of natural daylight.
º Second, we can assume the light intensity following a parabola, with area 2/3 of its base length, times the peak height: Lh2/3. To calculate the percentage of peak illumination corresponding to continuous illumination, we set peak illumination to unity = 1, and the length of illumination period will be the same. Hence, we equate 1L2/3 = L2/3 = hL, and solve for h = L2/3L= 2/3 = 0.67. Accordingly, the equivalent continuous illumination is 67 percent of peak illumination.
º Third, we assume the light intensity follows a Gaussian curve with area Π –2. Because a Gaussian curve has infinite left and right boundaries, we limit them to contain 95 percent of the area. This limit locates them at ±1.94 units from the center; for ease of calculations, we round that to ±2. We can equate that area to a rectangle with length 4 (from 2 to +2 underneath the Gaussian curve) and unknown height = intensity h, containing area 4h. Accordingly, we equate 0.95 Π = 4h and solve for h = 0.95 Π ^{2}2/4 = 0.42. The light intensity from a constant light source should be about 42 percent, or about 2/5 of the peak intensity of natural daylight.
These three values provide a range of sensible, constant light intensities of 42–79 percent of peak natural light intensity. Note that the natural light intensity curve has a shape in between the Gaussian curve and the parabola, which agrees with my personal experience of about half peak intensity or less for terrarium illumination intensity being appropriate.
— Daniel L. Geiger, PhD, is a marine invertebrate systematist working at the Santa Barbara Museum of Natural History. He is a member of the Orchid Society of Santa Barbara and the California Native Plant Society, and has been growing native California orchids outdoors, and more recently added an indoor terrarium with tropical orchids. Santa Barbara Museum of Natural History – Invertebrate Zoology, 2553 Puesta del Sol Road, Santa Barbara, California 93105 (email Geiger@vetigastropoda.com).
