Reefkeepers FAQ's

Part 1

The source of the Reefkeepers FAQ's is available here.

1.0 Water

1.1 Source Water - City Mains Water Is Not Good Enough

1.1.1 Background

U.S. EPA requirements for water quality from municipal sources are insufficiently pure for reef tank usage. For instance, the EPA standard for Nitrate (as NO3-N) is 10.0 mg/l, over twice the recommended maximum level. Extremely toxic (to inverts) heavy metals such as copper are allowed at levels as high as 1 mg/l.

Most public water supplies have contaminants well below the EPA levels and some reef tanks have done fine on some public supplies. In general, however, it is recommended that some form of post processing be performed on public water before it is introduced into the reef tank.

Although some people have access to distilled, de-ionized or reverse osmosis water from public sources, most will use a home sized system to produce their tank water. The two most common systems used are de-ionization resins, and reverse osmosis membranes.

1.1.2 DI filters

De-ionization (DI) units come in two basic varieties: mixed bed and separate bed. Two chambers are used in separate bed units, one for anion resins (to filter negatively charged ions), the other for cation resins (to filter positively charged ions). Mixed bed units use a single chamber with a mix of anion and cation resins.

DI units are 100% water efficient with no waste water. They are typically rated in terms of grains of capacity (a grain is 0.065 grams). Once the capacity of the unit is reached it either needs to be replaced or recharged (using strong acids and bases). Recharging is normally only an option for separate bed units.

A quick check of the local water quality reports (normally available free from the water supply company) will reveal the water purification capacity of a given DI unit. For example, if a unit rated at 1000 grains is purchased and the local water supply has a hardness of 123 mg/l (Missouri River, USA), then the unit capacity is (1000*0.065)/0.123 = 528 liters = 139.5 gallons of purified water.

Water production rates for DI units varies, but is typically around 10-15 gallons/hour.

Note that some contaminants captured by a DI unit may "break through" long before the unit indicates its capacity has been reached. Silica is a classic example. What happens is that silica is loosely bound to the resins initially, but is replaced by stronger binding materials like carbonates as the resins become exhausted. The use of two DI units in tandem, as mentioned elsewhere in this FAQ, helps to eliminate this problem.

1.1.3 RO Filters

Reverse osmosis (RO) units are normally based upon one of two membrane technologies: cellulose triacetate (CTA) and thin film composite (TFC). CTA based systems are typically cheaper and do not filter as well (90-95% rejection rates). TFC based systems cost more but have higher pollutant rejection rates (95%-98%). CTA membranes break down over time due to bacterial attack whereas TFC membranes are more or less impervious to this. CTA units are not recommended for reef tank purposes. TFC membranes are very sensitive chemically to the chlorine found in most water supplies. It is therefore very important to regularly replace the carbon block pre-filter associated with all better-grade TFC systems. TFC membranes are damaged by chlorine so a properly functioning GAC prefilter is mandatory.

RO filters work by forcing water under pressure against the membrane. The membrane allows the small water molecules to pass through while rejecting most of the larger contaminants.

RO units waste a lot of water. The membrane usually has 4-6 times as much water passing by it as it allows though. Unfortunately, the more water wasted, the better the membrane usually is at rejecting pollutants. Also, higher waste water flows are usually associated with longer membrane life. What this means in practice is that 300 gallons of total water may be required to produce 50 gallons of purified water.

Like any filter, RO membranes will eventually clog and need to be replaced. Replacement membranes cost around $50-$100. Prefilters are often placed in front of the membrane to help lengthen the lifetime. These filters commonly consist of a micron sediment filter and a carbon block filter. The micron filter removes large particles and the carbon filter removes chlorine, large organic molecules and some heavy metals. Of course, the use of prefilters makes initial unit cost more expensive but they should pay for themselves in longer membrane life.

RO units are rated in terms of gallons per day of output with 10-50 gallon/day units typically available. Note that the waste water produced by a RO unit is fine for hard water loving freshwater fish such as Rift Lake cichlids. Some route the reject water to the family garden.

The Spectapure brand of RO units has a good reputation.

1.1.4 Further Comments About Water

The ultimate in home water purification comes from combining the two technologies and processing the water from an RO unit though a DI unit. If a very high grade DI unit is used, water equivalent to triple distillation purification levels can be achieved. Since the water entering the DI unit can be 50 times purer than tapwater, the DI unit can process 50 times as much before the resins are exhausted. This significantly reduces the replacement or recharging cost of the DI unit. Using two DI units in tandem, moving the 2nd in as a replacement for an exhausted 1st unit, and replacing the 2nd unit with a new unit will insure that no undesirable elements "break through" the exhausted 1st unit and enter your supply.

If only one filter can be afforded, and waste water is not a concern, then it is recommended that a TFC RO unit with pre-filters be purchased. If waste water is a concern, or if only a small quantity of make-up water will be required (say, for a single 20 gallon tank), then a DI unit would be the preferred choice.

City water is unstable. Many cities modify their treatment process several times a year, dramatically changing its suitability for reef usage. For instance, Portland has great reef water - most, but not all, of the year.

1.2 Additives

Calcium (Ca)
required addition. A range of 400-450 ppm Ca++ (10-11 mM) is recommended. The preferred method is the usage of Kalkwasser (Limewater) for all evaporation make-up water. The use of Calcium Chloride (CaCl2) is known to cause problems with alkalinity (provable by balancing the relavent chemical reactions occuring in the tank when CaCl2 is added) and contributes to a shift in the ionic balance of the water which must be corrected via water changes. Still, CaCl2 is occassionally useful to repair serious Ca++ deficits.

The use of Calcium Carbonate reactors is growing in popularity as a replacement, or sometimes in addition to, Kalkwasser usage. Such reactors use CO2 recirculated through a bed of calcium carbonate (typically crushed coral like Geo-Marine) to reduce the pH and dissolve the calcium carbonate. A small fraction of the recirculated water is allowed to re-enter the aquarium and is replaced in the reactor with fresh tank water on a continuous basis. These systems are considerably more mechanically complex than Kalkwasser systems, often involving CO2 tanks, electrical valves, pH controllers, bubble counters, circulation pumps and related equipment. Once setup and tuned for the calcium demand in a tank, they can often be left alone for months. This low maintenance requirement is a primary benefit of the system, combined with the potential to inject more calcium into the system than kalkwasser alone could do due to kalkwasser's low solubility in make-up water. Note that calcium reactors may add residual CO2 to the system which can fuel algae growth. This extra dissolved CO2 may be purged by either enhanced gas exchange or by adding a small amount of kalkwasser to scavenge the CO2. Newer reactors often have a second carbonate stage to process and utilize this extra CO2. The efficiency of such a second stage is unknown to this author.

A third approach to calcium addition is the use of the newly available two-part ionically balanced solutions. These solutions use CaCl2 as a calcium source, but combine that material via the 2nd part solution with complex ion formulas that negate the normal problems associated with CaCl2 usage. These two-part solutions are relatively expensive for large aquariums but may be cost-effective for smaller tanks relative to the capital required for a calcium carbonate reactor. Others simply find their usage more convenient than the other alternatives. These additives will slowly raise the salinity of the water as a side effect. As always, monitor and correct your salinity as necessary.

Chelated calcium:

The efficacy of chelated calcium products available for reef aquaria is questionable. To the best of our knowledge, there exists no scientific evidence indicating that chelated calcium is especially available to corals and other CaCO3 depositing invertebrates. Nothing is known about the uptake of chelated calcium products by coral. And most importantly, there exists no evidence showing that chelated calcium products support stony coral growth rates in excess of, or even comparable to growth rates documented in aquaria where calcium is supplied as aqueous Ca(OH)2 [kalkwasser.]

Chelated calcium products also interfere with the ability to measure actual calcium levels in the aquarium. In particular, chelated calcium cannot be measured by any kit which uses EDTA titration, including the highly recommended HACH kit.

Until such a time as vendors supplying chelated calcium products make available well conceived, carefully documented uptake and growth studies with their products, or the same experiments are performed and published by third parties, we regard the use of chelated calcium products in the reef aquarium to be experimental at best, especially when kalkwasser and other non-chelated calcium sources are known to us to support the growth and even reproduction of stony corals in the home aquarium.

Iodine (I)
SeaChem and Salifert have recently introduced test kits which are finally allowing a view of actual usage in a reef tank. Note that iodine is naturally present in ocean water at relatively low levels (around 0.06 ppm - yes that 60 parts per billion). It is currently considered important for both soft coral growth and hard coral health. It is removed via skimming, activated carbon usage, and assimilation into biomass. It may also be removed by unidentified processes like precipitation.
Strontium (Sr)
used rapidly by most hard corals (weekly additions usually performed). Test kits are becoming available but the accuracy of current kits is still questionable. Natural ocean water levels of strontium is around 8 ppm.
increase alkalinity and control pH. Desired range is 2.5-3.5 meq/L (7-10 dKH) alkalinity. Alkalinity can be raised by the addition of one of many commercial buffer compounds. The addition of kalkwasser (saturated Ca(OH)2 solution - also known as "limewater"), which is often done to maintain calcium levels, will help maintain the alkalinity level. SeaChem's Marine Buffer, Reef Builder and Kent's Superbuffer dKH are popular. The Coralife and Thiel buffer products have had less favorable reviews.
Iron (Fe)
Used by alga. Add this if you want good macroalgae growth. Be sure that macroalgae growth is favored or else plague levels of hair algae may result.
Copper (Cu)
Used as a medication in fish-only tanks. Copper is highly toxic to invertebrates, even in very small concentrations. DO NOT USE THIS, IN ANY FORM, EVER, IN A REEF TANK OR ANY TANK WHICH CONTAINS INVERTEBRATES. PERIOD!
Other additives, especially the commercial "secret formula" mixtures, are more controversial. Some people report good results from some of them other people report disaster or no effect. from some of them other people report disaster or no effect. Experiment cautiously with them if desired. We recommend that all products that refuse to reveal what's in them be used with a great deal of caution.

1.3 Testable Parameters

Note: parts per million (ppm) and milligrams per liter (mg/l) are virtually identical in seawater and the units are used synonymously in this document.

1.3.1 Alkalinity

Alkalinity is a measure of the acid buffering capacity of a solution. That is, it is a measure of the ability of a solution to resist a decrease in pH when acids are added. Since acids are normally produced by the biological action of the reef tank contents, alkalinity in a closed system has a natural tendency to go down. Additives are used to keep it at a proper level.

Correct alkalinity levels allow hard corals and coralline algae to properly secrete new skeletal material. When alkalinity levels drop, the carbonate ions needed are not available and the process slows or stops.

Alkalinity is measured in one of three units: milliequivalents per liter (meq/l), German degrees of hardness (dKH) or parts per million of calcium carbonate (ppm CaCO3). Any of the units may be employed but dKH is most commonly used in the aquarium hobby and meq/l is used exclusively in modern scientific literature. The conversion for the three units is:

1 meq/l = 2.8 dKH = 50 ppm CaCO3

[As an aside, there is an imperial unit of alkalinity and hardness which is 'grains per gallon'. The water softening industry uses this unit. 1 gpg = 17 ppm CaCO3.]

A word of caution about the ppm CaCO3 unit is in order. The 'ppm CaCO3' unit reports the concentration of CaCO3 in pure water that would provide the same buffering capacity as the water sample in question. This does not mean the sample contains that much CaCO3. In fact, it tells you nothing about how much of the buffering is due to carbonates, it is only a measure of equivalency.

Alkalinity is often confused with carbonate hardness since both participate in acid neutralization and test kits may express both in either of the three units. However, carbonate hardness is technically a measure of only the carbonate species in equilibria whereas alkalinity measures the total acid binding ions present which may include sulfates, hydroxides, borates and others in addition to carbonates. In natural seawater, though, carbonates make up 96% of the alkalinity so equating alkalinity with carbonate hardness isn't too far off. As long as you're using a salt mix which yields an ion mix close to that of Natural Sea Water (NSW) you can also make this assumption. Some salt manufacturers alter the alkalinity component of their mix to increase the percentage of borates to (bi)carbonates in order to maintain a stabler pH in the aquarium. We do not feel this is good, and highly recommend you watch the trade magizines for reports on borates in salt mixes. (OK, OK, here's a preview... Instant Ocean does NOT have abnormal borates based on initial testing.)

Recommended values for alkalinity vary depending on who's work you read. Natural surface seawater has an alkalinity of about 2.4 meq/l. Following are levels recommended by various authors.

From John Tullock (1991) "The Reef Tank Owner's Manual":
page 46 - Alkalinity range should be 3.5 to 5.0 meq/l.
page 94 - Alkalinity reading of 2.5-5.0 meq/l is proper.
page 188- Alkalinity should be about 3.5 meq/l.
(In reference to maintaining Tridacna clams.)

Albert Thiel (1989), in "Small Reef Aquarium Basics" recommends 5.35-6.45 meq/l. This is an artificially high level which may initiate a "snowstorm" of CaCO3 precipitate. Most reef aquarists do not believe in such extreme and unnatural levels and recommend 3.0-3.5 meq/l as a good range instead.

The chemistry of how alkalinity, pH, CO2, carbonate, bicarbonate, and other ions interrelate is fairly complex and is beyond the scope and detail of this document.

Some recommended test kits for alkalinity are the SeaTest kit, the inexpensive Tetra kit and the LaMotte kit. The SeaTest kit measures in division of 0.5 meq/l or, if the amount of solution is doubled, 0.25 meq/l. The SeaTest kit uses titration in which the acid and indicator are included in the same reagent. The LaMotte kit is a little more expensive, though still fairly cheap, and is somewhat more accurate. The unit of titration is 4 ppm CaCO3 although in practice, one drop from the titration tube may be up to twice this amount making the resolution about 0.15 meq/l. The Lamotte kit has a separate indicator tablet and acid reagent which is a nice feature.

1.3.2 Calcium

Calcium content is referred to as 'calcium hardness' and is measured either in parts per million of calcium ion (ppm Ca++) or parts per million equivalent calcium carbonate (ppm CaCO3). Calcium hardness is often confused with alkalinity and carbonate hardness since the 'ppm CaCO3' unit may be used for all three. As with alkalinity, a calcium level expressed as X ppm CaCO3 does not imply that X ppm of calcium carbonate is present in the tank; it merely states that the sample contains an equivalent amount of calcium as if X ppm of CaCO3 were added to pure water. The reading also does not tell you how much carbonate is present.

Calcium hardness test kits are different from alkalinity kits. Some people have reported difficulties with the LaMotte calcium hardness kit. The Hach 'Total Hardness and Calcium' kit has not had these reports. Both express results in ppm CaCO3. The relationship between CaCO3 and Ca++ is:

1 ppm CaCO3 = 0.4 ppm Ca++

The results from a test kit reading in ppm CaCO3 may be converted to the molar concentration scale by dividing by 100.

100 ppm CaCO3 = 1 mM Ca++
40 ppm Ca++ = 1 mM Ca++

Calcium levels of natural surface seawater are around 420 ppm Ca++ (10.5 mM). In a well running reef tank you will notice, sometimes dramatic, calcium depletion. Calcium addition in some form is essential. A calcium level above 400 ppm is required and a range of 400-450 ppm Ca++ is recommended. Most reefkeeping books (see bibliography) explain the options for calcium addition.

1.3.3 pH

The suggested reef tank range is 8.3 to 8.4. The pH should hold its own unless alkalinity is low. If alkalinity is OK but pH is low there is probably a buildup of organic acids or a serious lack of gas exchange resulting in the retention/accumulation of CO2 which lowers pH.

Note that it is perfectly normal for the pH of a tank to swing considerably. There is a daily pH cycle where the pH is lowest just after the end of the dark period and highest sometime before the end of the light period. Having a pH range from 7.9 to 8.4 is not unheard of. Larger swings are probably indicative of low buffer levels or poor gas exchange.

1.3.4 Nitrate (NO3)

Two units are used to measure nitrates: nitrate (NO3-) and nitrate nitrogen (NO3-N or just N). The ratio is:

1 ppm NO3-N = 4.4 ppm NO3-.

Nitrates themselves may not be a problem but serve as an easily measured indicator of general water quality. Many hard to test for compounds like dissolved organics tend to have levels that correlate well with nitrate levels in typical tanks.

Different authors cite varying upper nitrate values permissible. No higher than 5 ppm NO3- is a good number with less than 0.25 ppm recommended. Unpolluted seawater has nitrate values below detectable levels of hobbyist test kits, so "unmeasurable" is the goal to strive for.

Most test kits measure nitrate-nitrogen. Do not forget to multiply by 4.4 to get the ionic nitrate reading. LaMotte makes a nitrate test kit that will measure down to 0.25 ppm NO3-N. The Hach kit, which measures down to 0.02 ppm N03-N has basic chemistry problems in saltwater and is no longer recommended.

1.3.5 Phosphate (PO4)

Phosphates, along with nitrates, are a primary nutrient of algae. Tanks with "high" levels of phosphates tend to be infested with hair algae. All authors cite zero ppm PO4 as a good goal. An upper level 0.1 ppm is recommended by Tullock (1991) with less than 0.05 ppm given by Thiel (1991).

The use of kalkwasser has been closely tied with reduction in phosphate levels. This may be due to precipitation of the phosphates at the kalkwasser injection site, or, more likely, due to increased export via skimming due to the associated higher system pH.

1.3.6 Specific Gravity

Short form: Specific Gravity is temperature dependant. See the next table for a quick lookup of the recommended hydrometer readings. They are based upon our recommended S.G. of 1.025 at 60 degrees F.

Degrees F. Hydrometer reading.  
50 1.0255  
55 1.0252  
60 1.0250  
65 1.0246  
70 1.0240  
75 1.0233  
80 1.0226  
85 1.0218 (rather hot for most tanks)
90 1.0210 (very hot for most tanks)

In more detail: 1.025 recommended for reef tanks. Note that virtually all hydrometers are calibrated for measurements at a temperature of 60 F. Included below is a short table of temperature adjustments. Add the value shown to your hydrometer reading to get an accurate reading.

Degrees F. Correction
50 -0.0005
55 -0.0002
60 0.0000
65 0.0004
70 0.0010
75 0.0017
80 0.0024
85 0.0032
90 0.0040

For example: If the hydrometer reads 1.0235 at 80F, the actual Specific Gravity is 1.0235 + 0.0024 = 1.0259

Note: If your tank is between 75F and 80F, this means you should try and keep your Specific Gravity around 1.0230 to 1.0235.

For all practical purposes, the scale is linear between data points, so you can simply extrapolate between table entries. For instance, 78F is 3/5 the distance between 75F and 80F; the difference in corrections is 0.0024-0.0017 = 0.0007. 3/5th of 0.0007 is 0.0004. Add the offset 0.0004 to the base value for 75F of 0.0017 and you get a correction value for 78F of 0.0021.

It is fairly common in literature to see references to salinity in terms of Parts Per Thousand (PPT). For salinities in the range we are interested in, the conversion formulas are:

Salinity = 1.1 + 1300 * (Temperature corrected Specific Gravity - 0.999)
Temperature corrected Specific Gravity = ((Salinity - 1.1) / 1300) + 0.999

Here is a short table of some common values:

Salinity Specific Gravity  
20 PPT 1.0135  
25 PPT 1.0174  
30 PPT 1.0212  
35 PPT 1.0251 * Typical Ocean Value *
40 PPT 1.0289  

1.4 Water Changes

"The solution to pollution is dilution". Water changes are used to correct problems. Minimal changes of 5%/year when all is set up and running smoothly may suffice. Some feel that an occasional water change of about 20% every 1-3 month is a reasonable safety net that may help prevent contaminant buildup, shift in ion balance, and trace element depletion problems. Others recommend 5%-10% per week.

2.0 Filtration and Equipment

2.1 Live Rock

Live rock is simply old reef substrate that has become the home to multiple small plants and animals. Pieces vary in size and shape from baseball size to dinner plate size in typical tanks. In large tanks (> 500 gallons) very large pieces of live rock tend to be used. These pieces may individually weight up to 85lbs (about the limit of what one person can handle).

The use of live rock greatly increases the bio-diversity in a tank. However, its primary purpose is to provide a home for bacteria that provide the biological filtration for the aquarium.

Cheap rock has low amounts of coralline algae and tends to grow hair algae well. It may be suitable for a soft coral only tank. Hair algae free coralline encrusted live rock (high quality Florida and/or pacific (Marshall and Tonga Island) rock is highly desirable. "Berlin" style tanks use high quality live rock (and protein skimming) as the primary filtration method with great success.

Although an old rule of thumb states that 1-2 lbs of live rock is required per gallon of tank size, the wide range of available rock makes the rule pretty inaccurate. It is suggested that a visual method be used, consuming approximately 1/3rd of the tank volume with rock - leaving 2/3rd of the volume in open water. You should probably only use the rule of thumb as a sanity check. For instance, 10 lbs of the best rock would be too little for a 75 gallon tank, no matter how good the rock is. Likewise, 300 lbs would be overkill.

Live rock is typically "cured" prior to introducing other life forms in a tank. This curing process is, in its shortest form, simply a period of time to allow dead and dying organisms on the rock a chance to decay. Any time live rock is moved, some organisms will probably die. Shipping rock submerged in oxygenated water (very expensive) is the only practical way to minimize this die-off.

Live rock should be cured in a container with excellent protein skimming, activated carbon, excellent oxygenation and water motion. There is a very real danger of anoxia when freshly shipped live rock is placed into a curing vessel. Unless the dissolved oxygen concentration is kept high in the curing vessel, the bacterial bloom fed by the initial die-off will cause the curing vessel to become anoxic, and even more life will perish. Therefore gas exchange and water motion are crucial. Protein skimming helps to remove organics before they are consumed by bacteria. Addition of a cycled biological filter may reduce the severity and lethality of the ammonia spike when curing live rock.

It is recommended that fresh live rock be allowed to cure for at least one month prior to the introduction of any other life forms in a newly setup aquarium. There may well be advantages to waiting between three months and a year, with the tank running in normal mode (full circulation, heating, lighting, etc), before adding other life forms in order to allow the biodiversity naturally present on the rock to stabilize.

2.2 Protein Skimmers

Protein skimmers are devices that mix large volumes of air and tank water to produce foam. This foam is then collected and disposed of. The foam will contain a fair amount of particulate material, lots of organic material that would otherwise breakdown and pollute the tank, and unfortunately some trace elements like iodine. Besides removing organic material, skimmers play a key role in maintaining proper O2 and CO2 levels in tank water. For instance, having a large skimmer that processes a large amount of air will allow larger quantities of kalkwasser addition, both due to increased evaporation, and due to improved CO2 absorbtion capacity.

Required equipment. Don't undersize. Common wisdom is that you can't overskim a tank. Recent developments in using down-draft style skimmers, with ETS being the first commercial instance, have raised the possibility that it's now possible to overskim a tank. This is stated with a lot of caution, we still feel that its impossible to overskim using airstone or venturi driven skimmers of reasonable size. (Using a 8" x 6' counter-current skimmer processing 600 gph of air on a 20 gallon tank could overskim it - be reasonable!)

Unfortunatly, there is no formula to determine the required size of a skimmer. Amount of organic waste generating organisms (fish, coral, live rock, etc.) will obviously be the primary variable. All skimmers should be filled with tiny bubbles and have a milky white appearance. Any skimmer that doesn't match that requirement is not working optimally.

There are some basic rules-of-thumb on minimum skimmer sizing however: A skimmer should process at least one tankful of air and one tankful of water per hour. For most tanks, the water rate is easy. The air rate however is not. Most counter current (explained below) skimmers are under-supplied with air. If you have a sealed skimmer where the air can only exit from one fitting, its easy to measure the flow rate. Simply take a large plastic bag (something in the 2 gallon size works well), empty it, and place it over the air exhaust port. Time how long it takes to fill and do the math.

Three basic styles of skimmers exist: counter current air driven, venturi driven, and down-draft. All styles work fine, all have tradeoffs. All require some tuning. Expect to spend some time over the first month or so learning how to keep your skimmer tuned.

2.2.1 Counter Current Air Driven Protein Skimmer

These skimmers usually require three pieces of equipment typically not sold with them: an air pump, air stones and a water pump. Total skimmer cost depends upon the kinds of equipment needed to run the skimmer properly.

The water pump injects the water to be skimmed into the unit. Some people use gravity to feed surface overflow water to the skimmer or divert part of the main circulation pump's return flow into the skimmer to eliminate the need for a dedicated pump. Otherwise a powerhead in the sump usually suffices for the water pump. Some arrangement should be made to gather surface water for all forms of skimmers.

The air pump must be large enough and a sufficient number of air stones must be driven to make the skimming column milky white. In some smaller skimmers one medium sized air pump like a Tetra Luft G and one air stone will be sufficient. Other skimmers need a lot more to perform optimally.

Speaking of air-pumps, we find it baffling that folks who would consider spending $250 for a water pump to drive a different style skimmer totally reject spending anything on that order for an air-pump. When comparing skimmers, please be reasonable. Don't expect a counter-current airstone skimmer driven by a Tetra Luft to function as well as a venturi or down-draft skimmer driven by an Iwaki RLT-75 water pump costing around 10 times as much.

Air driven skimmers should use limewood air stones which will need to be replaced from time to time. Cheap limewood air stones have a reputation of needing to be replaced much more often than high quality stones. Coralife limewood air stones have a good reputation. Air stone replacement rate depends on your tank and skimmer; some people need to change them every 2 weeks others only after 3-4 months. It is believed that having a high air-flow prolongs the life of airstones. Some folks are experimenting with using VERY high quality fine-pore ceramic airstones, such as those available from Aquatic Eco-Systems.

A.J. Nilsen recommends a 1x tank volume per hour turnover of both water and air by counter current air driven skimmers. Others feel each skimmer has an optimal rate of air and water processing and that if more skimming is desired then more or bigger skimmers should be added rather than trying to operate the current one beyond its optimal performance range.

Some hold that any skimmer under 4' high and 4" in diameter is too small for anything over about a 20 gallon reef.

2.2.2 Venturi Protein Skimmers

These skimmers use the Bernoulli effect of the venturi valve to inject air bubbles into the water. This obviates the need of an air pump and air stones. The penalty is that a relatively large, high pressure (read expensive and power hungry) dedicated water pump is mandatory for the venturi unit to inject sufficient amounts of air.

A particular commercial venturi skimmer may or may not come with a water pump. If it does supply a pump, it may or may not be sufficiently large to run the skimmer properly. At least some of the venturi skimmers easily available are not very well designed.

Venturi valves require occasional cleaning of the air opening. This is as simple as reaming the opening out with a pipe cleaner every few days. An acid bath may be required if the unit clogs or gets coated with mineral deposits.

Most venturi style skimmers are more compact than CC skimmers. Manufacturers state that they are more efficient, since they (supposedly) inject more air. Many suspect that design constraints (back pressure severely affects venturi performance) have more to do with the manufactured height (who would want a top injected 4' skimmer with air only in the top foot of water?). Properly designed venturi skimmers are tall to maximize air contact time, and require pumps that can handle backpressure.

2.2.3 Down-draft Skimmers

This style skimmer was first commercially introduced as the ETS skimmer. Currently there are several on the market. This skimmer using a high pressure stream of water injected downward into a column of bio- balls to suck air into the water stream and break the air up into tiny bubbles. They do this very well, with some casual testing indicating air movement on the order of 10 times the volume of equivalently pumped venturi models.

The cost of ETS skimmers is relatively high, but expected to drop as competitors enter the market. Cost of operation is high due to the need for a huge water pump. The use of Iwaki 55s and 70s on base unit ETS skimmers is common. This size skimmer is appropriate for a 70-135 gallon tank. e.g. The skimmer pump may be larger than many use for the main tank circulation.

Construction of down-draft skimmers is easy, designing one to function optimally is not.

As previously mentioned, this is the first style skimmer where the authors of this FAQ have even considered the possibility of over- skimming a tank. With this style skimmer, bigger may actually not be better...

2.2.4 Protein Skimmer Considerations

Below are some pros and cons of three styles of skimmers. Some people will debate some of the statements.

Venturi skimmers, due to the large water pump needed, have a higher initial purchase price than CC units for the same amount of skimming. Many venturi skimmers are poorly designed, with woefully inadequate pumps to drive the venturi valve. Remember: Its not the technology that makes a skimmer good, its the amount of air and water processed that makes it good. The technology is just a method to reach that goal. There are plenty of counter current skimmers that out-perform venturi based skimmers, and visa-versa.

The operational cost of a venturi unit is basically just the electricity bill. A CC unit must sum in electricity consumption for the water pump and air pump (usually small) plus air stone and diaphragm replacement. Which one is more cost effective for you depends upon which equipment you had to buy to run the skimmer properly, your electricity rate and how often air stones need to be replaced. Most people find CC skimmers less expensive to both purchase and operate for the same amount of skimming.

Venturi skimmers are less cumbersome in appearance and in operation. They are usually smaller and quieter. They are on the whole more hassle free. The powerful pump required for venturi skimmers may, however, add considerable heat to the water.

When large down-draft skimmers are driven by an appropriate pump, they outperform venturi skimmers that use pumps of similar power consumption. That is, they are more "efficient". Once setup and past their 3-14 day break-in period, down-draft skimmers require no maintanence beyond the periodic cleaning that all skimmers require.

There is some debate over down-draft skimmers vs properly run counter-current airstone skimmers. A counter current skimmer can be provided with a similar air and water flow for somewhat less initial money. Note that we are not comparing $1200 ETS systems to $30 Coralife co-current skimmers here, but rather systems about the same physical size.

Of the three skimmer designs, counter current skimmers are the most plankton friendly. Although some small plankton will be removed as particulate material, it is at least conceivable for plankton to survive the trip through the water pump and through the bubble column. Venturi skimmers are much harder on plankton, since significant pressure is applied to them as they pass through the venturi valve. Down draft skimmers are assumed to kill anything larger than single cell organisms that pass through them due to the force and mechanical stress the water is exposed to.

One general note on water pumps: The amount of heat added to the water varies by brand, design, usage, and placement. Basically, the more efficient the pump (gallons delivered at a given pressure for a given power usage), the cooler it will run. Restricting the output of the pump will generally increase the water temperature. (Never restrict the intake of a centrifugal pump!) Obviously, an air cooled pump will increase your tank temperature less than a submersible (and therefore tank water cooled) pump will.

2.3 Granular Activated Carbon

Some debate about its usage. Most use it at least a few days a month, some continuously. Many brands have problems with phosphate leaching.

GAC has the ability to very rapidly remove dissolved organic compounds which cause the water to yellow. Indeed, failure to remove these compounds is a excellent way to determine when your carbon has been exhausted. A simple test consist of collecting (even temporarily) 5 gallons or so of tank water in a white plastic container. If the water appears yellowish, the carbon should be replaced.

WARNING: If the tanks water is significantly yellow, carbon should be replaced very slowly, like a gram-per-gallon at a time. Failure to do this may drastically improve the water's clarity, allowing more UV light to reach the organisms. Corals have been known to bleach and die after large carbon changes due to this rapid light transmission change.

2.4 Other Chemical Filter Media

X-Nitrate, X-Phosphate, Polyfilters, Chemi-pure, etc. - probably not needed in an established, balanced reef aquarium. A prominent manufacturer of these materials was either unwilling or unable to supply capacities for removing the named compounds from seawater. They may cause adverse reactions in some inverts.

2.5 Mechanical filtration

This is an area of interest currently being debated. Originally the FAQ stated:
Good idea to pre-filter skimmer water. Floss works fine and is cheap and disposable. Sponges work well, but require cleaning twice a week or so. Natural sponges with a medium fine or fine pore size are recommended. Some people don't use mechanical filtration, allowing detritus to settle in places for removal by siphoning. Some of these people make dedicated "settling tanks" to trap debris in a convenient place.
Julian Sprung suggests not pre-filtering skimmer water as skimmers will remove particulates (rather than trapping them as a pre-filter would do). Spotte confirms this and terms this filtering mechanism as 'froth flotation'.

Many members of the group of authors do not use mechanical filtration. They believe that such systems filter out the plankton that is used as food by many marine organisms. Some members use "live sand" setups, with detritivores. Others routinely siphon accumulated detritus.

Use of a mechanical filter for short periods may help when attempting to resolve specific problems, such as a hair algae outbreak.

2.6 Under Gravel Filters (UGF)

Not appropriate for a Reef Tank. Although they will work for 6 months or so, eventually detritus buildup will cause a nitrate problem. Long term, it's virtually impossible to keep nitrates below about 40 ppm NO3- which is way too high for corals.

Remember, NO3- is an indicator for other waste compounds (e.g. dissolved organics) which are not easily measureable and these compounds will also be present in abundance.

2.7 Reverse Flow UGFs

An attempt to solve the detritus buildup problem associated with normal flow UGFs. It's a good idea that doesn't work well in practice. This system has problems with uneven water flow due to channeling within the bottom gravel.

2.8 Trickle Filters

Also known as Wet/Dry Filters. An improvement over UGF and RUGF filters. Nitrates can be kept low (say, around 5 ppm) with adequate water changes. It does not seem to be possible to keep nitrates very low (less than 1 ppm) if a trickle filter is the sole biological filtration. Those that report less than 1 ppm normally have adequate live rock, and find that their nitrates remain low even (and often get lower) when they remove all the bio-material from their trickle filters (turning them into plain sumps, useful for holding carbon and as a water reservoir).

2.9 Algae Scrubbers (long)

Summary: The jury is still out, even after several years of experience. Some species of corals seem to do fine, and occassionaly even better, in algae scrubed tanks. Other species simply die. In any case, we still recommend this only for experienced aquarists who fully understand the risk they are taking. What's included below is a decent summary of how folks felt a couple of years ago. Most have softened their view abit since then. Few consider the systems the ultimate solution...

In most healthy natural communities, particularly coral reefs, dissolved nutrients are scarce. In aquaria, by contrast, nutrients in the form of dissolved inorganic nitrogen, or DIN, (a collective term for ammonia, nitrites, and nitrates) accumulate very rapidly as fish and other organisms excrete these wastes. The most basic problem in any aquarium is limiting the accumulation of DIN.

In reef aquaria, DIN is consumed by the community of organisms on the live rock. It is uncertain what relative contribution is made by bacteria as opposed to algae, but it is certain that the live rock community as a whole can remove a substantial amount of DIN from a reef aquarium. In fact, it is quite possible to run a reef tank with no biological filtration (DIN consumption) other than that which takes place on the rock. This method is part of what is now known in the United States as the "Berlin school" of reefkeeping.

Other schools of thought utilize additional biological filtration in separate filters. Traditional reef tanks supplement the filtration provided by the reef (often not acknowledging the role of the reef itself) with bacteria-based trickle filters. Many readers probably learned this technique first, as it has been the dominant method in the United States amateur hobby for some time. Yet another approach uses algae, which are also capable of utilizing inorganic nitrogen directly. An algae filter, or algal scrubber as it is usually called, is simply a biological filter which utilizes a colony of algae rather than bacteria as consumers of inorganic nitrogen.

Algal scrubbers are not new; they are discussed in Martin Moe's (1989) excellent Marine Aquarium Reference: Systems and Invertebrates, for example. However, algae filters have been regarded in the past as too bulky and inefficient to be the sole filter for a aquarium. The recent surge of interest in algal scrubbers seems to have been generated by Adey and Loveland's book Dynamic Aquaria (1991). They discuss both techniques which allow an algal scrubber to be compact and efficient and also a number of arguments as to why they are preferable to other filtration methods.

One reason to use an algal scrubber according to Adey and Loveland is that it mirrors the way DIN is cycled in nature. They claim that perhaps 70-90% of the DIN in reef communities is consumed by algae, rather than by bacteria. The two methods produce rather different water chemistry; for example, algae are net producers of oxygen and remove carbon dioxide, while a bacterial filter consumes oxygen and produces carbon dioxide. They argue that it should be easier to maintain the type of water chemistry found over a natural reef by relying on an algal scrubber.

Also, algae remove the nitrogen from the water in order to build tissue, while filter bacteria simply put it into a less toxic form. The excess nitrogen can be removed completely by periodic algae harvests, while dissolved nitrogen in the form of nitrate is not as easy to remove. Adey and Loveland claim that their methods can bring levels of DIN down to a few hundredths of a ppm, far below (in their opinion) the levels reachable with other methods. A related argument in favor of algal scrubbers is that stability in natural ecosystems comes from locking up nutrients in biomass, not in allowing it to be free in the environment. An algal scrubber does precisely this, while a bacterial filter converts it to free nitrate dissolved in the water.

A final reason to use an algal scrubber according to Adey and Loveland is that many other kinds of filtration (including protein skimmers) remove plankton from the water. An algal filter naturally does not do this, and can actually provide a refuge for some forms of plankton. The importance of this effect is, however, a matter of some debate.

As compelling as some find the above arguments in theory, there seem to be serious problems with algal scrubbing in practice. Many attempts by public aquaria at implementing reef tanks using only algal scrubbing have been failures. In particular, it seems difficult to find successful long term success with Scleractinia (stony corals) in such tanks, and those success stories which can be found are quite difficult to verify and often contradicted by others.

Various public and private aquaria have used algae scrubber filters on their reef aquaria, with disastrous results. The microcosm at the Smithsonain Institution has yet to keep scleractinia alive for more than a year. While Dr. Adey has stated how well corals grow in this system, those viewing the system have failed to find these corals. In an interview with Jill Johnson, one of the techs responsible for the Smithsonian tank, she stated to Frank M. Greco that frequent collecting trips were needed to keep the system stocked with live scleractinia.

The Pittsburgh AquaZoo also has a "reef" tank based on Dr. Adey's algal scrubbers. This tank is nothing more than a pile of rocks covered with filamentous green algae, and the water is quite yellow (as is the Smithsonian tank) from the presence of dissolved organics (ORP readings have been around 165). As with the Smithsonian tank, scleractinia do not survive longer than a few months. The same applies to soft corals as well. When I (Frank M. Greco) saw this tank on May 3, 1993, there were no living corals to be found even though a collecting trip to Belize was made several months earlier and 81 pieces of living scleractinia were brought back. There were, however, two piles of dead Atlantic scleractinia: one right behind the tank and the other in the greenhouse housing the algal scrubbers.

The Carnegie Science Museum (Pittsburgh, PA) also uses an algal scrubber system, but with significant modifications. This tank looks the best of the three. There are several species of hardy Scleractinia and soft corals that are doing quite well. The water is clear (a bit cloudy). The major differences between this system and the other two is the use of carbon, a small, barely functioning algal scrubber, about 1000 lbs. of excellent quality live rock (Florida), water changes, and the addition of Sr and Ca.

The last system I know of that uses an algal scrubber is the Great Barrier Reef Microcosm in Townsville, Australia. As of this writing, the system is not maintaining live Scleractinia, and frequent collecting trips are needed in order to replenish the exhibit. It should also be noted here that while Dr. Adey has claimed in his book Dynamic Aquaria that corals have spawned in this system, what he doesn't mention is that the corals which spawned were collected only months before the known spawning season. From these few examples, it should be clear that algal scrubbers are NOT to be used in systems containing live scleractinia.

[Some theories, observations, and other comments withdrawn.]

The weight of evidence at this point seems to be against the use of algal scrubbing in reef tanks, and the method should be considered to be experimental. Beginners particularly are advised to avoid this technique until they have considerably more experience with reefkeeping. The advanced aquarist may well wish to experiment with this interesting and controversial method, but it would be unwise to risk the lives of an entire reef tank full of coral. Such experiments should progress slowly, beginning with the most hardy of inhabitants. Many of the objections center on stony coral survival, and it is possible that scrubbed tanks with fish and hardy invertebrates may do quite well.

2.10 Live Sand

Of relatively recent interest in the hobby is the use of "live sand". Live sand consist of small grain (0.5mm-1.0mm) coral sand that is populated with crustaceans, worms, mollusks and bacteria. It is normally used at a rate of 10lbs per square foot of bottom area - which yields about a 1" deep covering. Variations from 1/8" to 3"s of covering have been reported.

If you decide to have a live sand substrate bottom, you should include several creatures that will turn-over, or otherwise, move the sand around. Recommendations include: Sea Cucumbers, Brittle Starfish, Serpent Starfish, Orange Spot and Golden Headed Sleeper Gobies, Yellow Jawfish, Watchman Gobies, and other detrivores. A mix of the above is recommended, since each creature moves the sand around differently.

If you use sea cucumbers make sure there is NO way one can enter into any pumps. If a cucumber gets stuck in a pump, it will potentially release extremly toxic substances into your tank. The only remedy is to start your reef tank over since no known anti-toxin exist. Yes, everything may die in your tank (strong skimming may save the day, but don't count on it).

Live sand has a reputation of eliminating the final traces of nitrates in otherwise well run tanks. It also provides an environment for additional bio-diversity in the tank. Additionally, some feel that the chemical balance and stability of a tank's water is improved when live sand is present.

Note that live sand usage should still be considered experimental. Usage is dependant upon have the sand sifted and otherwise moved around to prevent detritus from accumulating. Many people have reported problems keeping their turn-over creatures alive for long periods of time. Some have not seen the reported nitrate reductions. Keep in mind that many reef tanks have operated for years without a substrate and have no detectable nitrate concentrations. Use of very fine sand has been linked to hydrogen sulfide production in tanks. On the other hand, use of live sand definitely allows for a more diverse bio community.

3.0 Lights

3.1 General Discussion

A rough "rule of thumb" is 4 Watts/gallon with successful tanks using from 1.5 - 6 Watts/gallon.
  1. Fluorescent fine (some prefer) for shallow (<20") tanks. Use mix of bulbs (50-50, 03s, etc.)
  2. Metal Halide (MH) required for deeper tanks.
  3. Mercury Vapor, Halogen, HPS, etc. - avoid, wrong spectral output.

3.2 Detail Discussion

For most aquarium lighting applications, the bottom line is getting the needed intensity and spectrum of light at the lowest cost while remaining within aesthetic limits.

A lighting analysis is now presented. Everyone has their own sets of numbers they would plug in here, for now let's assume the following for comparison. Many will debate the specifics found below. Feel free to substitute your own numbers, but the methodology is sound.

Bulb cost and performance:

NO lumens per lamp = 2600 (Phillips F40D daylight, initial)
NO watts per lamp = 40 (ditto)
NO cost per lamp = ~$20 (from memory, DLS actinic day)

VHO lumens per lamp = 5940 (Phillips F48T12/D/VHO daylight, initial)
VHO watts per lamp = 110 (ditto)
VHO cost per lamp = ~$30 (ditto)

MH lumens per lamp = 36000 (Philips MH400/U, initial)
MH watts per lamp = 400 (ditto)
MH cost per lamp = ~$70 (from memory, Venture 5200K)

operate lamps 12 hours/day
replace lamps once per year
electricity cost = $.09 / KWH (your mileage may vary)

Annual cost per lumen:

cost = ( cost-per-lamp / lumens-per-lamp )
+ ( watts-per-lamp / lumens-per-lamp ) * 12 * 365 * .09 / 1000

NO cost = .0077 + .0061 = .0138 dollars per year per lumen
VHO cost = .0051 + .0073 = .0124 dollars per year per lumen
MH cost = .0019 + .0044 = .0063 dollars per year per lumen

Basically, in fluorescents, the VHO lamps give a higher operating cost but a lower replacement cost for the same total amount of light. But it's close, and you should plug in your own numbers to see what's best for you. If you replace lamps more frequently then VHO is better, if you pay more for power, NO is better.

There is a greater variety of lamps available for NO than VHO. OTOH, it seems that NO lamps can be operated at VHO power levels, with a somewhat shortened lifetime (the higher replacement frequency is offset by lower lamp cost), so this may not be an issue.

The initial installation cost (basically the ballast cost) is higher for VHO, even in terms of per-lumen, but this is a pretty small part of the total cost of the lighting system over the years.

NO requires more lamps for a given total light intensity, so you may not be able to fit enough NO bulbs in your hood if you need a lot of light.

MH seems to be a winner in both replacement and operating costs. The color spectrums available in MH lights has improved substantially recently with the advent of 10K and 20K bulbs. 10K bulbs are becoming very popular as the sole light source for reef tanks. 20K bulbs are often being used in conjunction with older lighting systems to provide a more balanced light mix.

On the flip side, MH bulb vendors have had some horrible quality problems, and obscene pricing practices. Recent testing has also determined that MH ballast have a manufacturing acceptable level of output variance that may result in totally unacceptable differences in individual bulb spectrum output.

MH vs fluorescent also gets into the aesthetic and biological considerations. Water surface ripples causing light ripples in the aquarium and room are pronounced with MH lighting. Many people appreciate this effect. Some (e.g. Julian Sprung) feel the variation in light intensity is actually important for some photosynthetic organisms.

Many people are under the impression MH runs hot, whereas fluorescent doesn't. In reality, the efficiencies are similar, with MH producing slightly LESS heat than the equivalent fluorescent. The difference is MH dumps all the heat in a small space so the local temperature rise is greater. But if you want to try to get rid of the heat it's actually easier to do it if the heat is concentrated in one spot, since its easier to get rid of a small amount of very hot air than a very large amount of warm air.

A separate issue is the selection of a conventional ballast vs an electronic one. There is no doubt the electronic ones are more expensive to purchase, but the savings in electricity offset the high initial cost in a year or so. Also, if heat production is an issue, the electronic ballasts are to be favored. The Icecap VHO electronic ballast is widely advertised, however its advertised claims are also frequently questioned. Advance makes a series of NO electronic ballasts.

There are yet two more issues, for which there are a lot of questions and too few answers. Specifically, the short term flicker in light intensity, and radiated electromagnetic fields.

Fluorescent lamps on conventional ballasts flicker at 120 Hz, which is above the human visual response, so we don't see it (actually, the flicker is both in intensity and spectrum). But that doesn't mean other creatures can't see it, or whether they benefit or are disadvantaged by it. Electronic ballasts cause flicker at ~30 KHz; it is seriously doubtful that any creature can detect this, so it would appear constant.

The flicker doesn't have to be visible to have an effect: it causes any movement to appear strobed, and this may affect the feeding efficiency of visual hunters.

The fields issue is even more obscure. At least many cartilaginous fish (sharks, rays, etc) are known to be extremely sensitive to electric fields, and many crustaceans are sensitive to magnetic fields (crabs with pieces of magnetite in internal sensory organs). Fluorescent lamps, with the large area they cover, tend to radiate (using the term pretty loosely) fairly strongly, but MH, and the wiring, and the ballasts can radiate too. It's unknown how significant this could be in an aquarium (but its known sharks preferentially attack undersea cables because of the fields, so there is at least indirect evidence its an issue worth some thought).

BTW, a grounding device reduces the level of induced voltages in the tank, but this is achieved at the expense of increased induced current, so its effect (if any) may depend on the species. Also, note if you have a titanium coil chiller on the tank, it is probably already grounded through the chiller, and an additional ground may in fact increase the electric current. This should not be an issue with epoxy or ceramic coated chiller coils.

4.0 Cost Estimates

Here is a rough estimate of what setting up a reef tank may cost. Two cases are included: a 20g micro-reef and a 70g mini-reef. The estimates show the min and max for most of the common pieces of equipment. The estimates assume a standard type of filtration that is popular today. If a different setup is used, the price could be more or less. The equipment includes a tank with some sort of siphon/drain to a sump and then a return pump back to the tank. A protein skimmer is installed in the sump. This setup is similar to a typical wet/dry trickle filter except there is no trickle section with media. This allows the use of simpler, less expensive sump although a commercial W/D without media could be used. A trickle media could be utilized at greater cost although many reefkeepers think it is unnecessary. Keep in mind that prices sometimes vary geographically. Also, availability may vary. For example, reasonable Florida live rock will soon no longer be available (at least not for $2-4/lb).

The estimates include the cost of the initial set-up. There is also a section on ongoing costs. The ongoing cost will vary greatly, especially considering that you will stock your tank gradually. Keep in mind that you always end up spending more than you think you will. If you set up a reef, you will end up stopping at the hardware store and/or aquarium store for timers, extension cords, GFIs (a must!), buckets, hoses, and books, don't forget books. You should read a few books on reefkeeping before even planning your setup. An extra hundred bucks or three is going to leak out of your wallet whether you plan on it or not.

Another factor is that more advanced equipment may translate into less or easier maintenance. You should keep in mind that if you go with inferior equipment, maintaining the tank will be more work. More expense will mean more automated equipment and less work. Also, some varieties of inverts require more exacting condition, more light, etc. Plan your purchases so that the stock you buy has a chance of surviving with the equipment you are using. If you have a bare minimum system, stick hardy items like soft-corals, polyps, mushrooms, etc.

The minimum included is close to rock-bottom as far as an acceptable systems goes. It assumes that you are DIYing much of the equipment as cheaply as possible. The maximum in the estimate is in some areas a little extravagant but not unreasonable. A good system that is not extravagant could be put together for somewhere in between the two extremes. Perhaps, for 1.25 to 2 times the minimum, you would have a very nice system. Some areas are easier to cut-corners on than others and some of the initial cost may be incremental, like buying test kits as needed. Also, you may have some of the equipment already from previous set-ups or be buying it used. Seek out the advice of an experienced reefkeeper when planning and pricing your system.

Micro-Reef (20g)

Tank $ 20/ 140 Glass/
Stand 0/ 250 Sturdy piece of furniture/
Fancy acrylic stand.
Lights 100/ 300 DIY 60W fluorescent/
70W or 150W MH hood or pendant.
Main Pump 20/ 60 Large powerhead/
Hobby pump.
Sump 10/ 120 A plastic storage container from the hardware store/
A small commercial W/D without media. (A nice DIY acrylic sump can be built for about $40.)
Skimmer 60/ 220 DIY skimmer, power head, air pump/
Small commercial venturi unit with integral pump.
Plumbing 30/ 100 DIY overflow and misc pipes, etc/
Drilled tank or commercial overflow box plus misc pipes, etc.
Live-Rock 140/ 400 35lb case of Fla rock plus shipping/
30lbs of Pacific rock plus shipping.
Water Treatment 100/ 600 DIY mixed-bed DI with carbon prefilter/
TFC RO unit with DI postfilter and automated top-off.
Test Kits 100/ 500 A SW combo kit plus and Alk and Ca test/
Most of the Lamotte and/or Hach kits you think you might need.
Salt 10/ 20 One 50g bag, price varies.
Accessories 20/ 200 There are a variety of gadgets you could get. You might want to start with a net or two and maybe a pair of tongs.
Setup Total $ 610/2910  

Mini-Reef (70g)

Tank $ 140/ 350 Glass/
Stand $ 100/ 500 Cheap wood or iron stand/
Fancy acrylic stand.
Lights $ 200/ 600 DIY 160W fluorescent/
2x150-175 MH hood (possibly with Actinics).
Main Pump $ 80/ 140 400-600gph, price varies with brand.
Sump $ 10/ 200 A plastic storage container from the hardware store/
A commercial W/D without media. A nice DIY acrylic sump can be built for about $50.
Skimmer $ 80/ 450 A DIY skimmer,powerhead,air pump/
A large commercial venturi unit with a large pump driving it.
Plumbing $ 50/ 150 DIY overflow and misc pipes, etc/
Drilled tank or commercial overflow box plus misc pipes, etc.
Live Rock $ 460/1200 140lbs Fla rock plus shipping/
110lbs Pacific rock plus shipping.
Water Treatment $ 100/ 600 DIY mixed-bed DI with carbon prefilter/
TFC RO unit with mixed-bed DI postfilter and automated top-off.
Test Kits $ 100/ 500 A SW combo kit plus and Alk and Ca test/
Most of the Lamotte and/or Hach kits you think you might need.
Salt $ 20/ 40 Two 50g bags, price varies.
Accessories $ 40/ 500 There are a variety of gadgets you could get. You might want to start with a net or two and maybe a pair of tongs. You could get wave-makers, circulation pumps and lots of other do-dads.
Chiller $ 0/ 600 Don't use a chiller, live somewhere cool, keep the tank in the basement, or an adequately air-conditioned room/
A commercial chiller.
Setup Total $1380/5830  

Ongoing Costs

Most reefkeepers believe that some additives are necessary. At minimum, a buffer compound is needed to maintain the alkalinity. Also, some Calcium supplement such as Kalkwasser or Cacl2 should be used. A few trace additives like Strontium and Iodine/Iodide should also be added. The initial supply of these products will be around $50. The ongoing rate will vary depending on the size of the tank.
Water Purifier
If you go with a DI system, you will have to replace and/or recharge resin. An RO system will require periodic replacement of the membrane and prefilters. In the long run, maintenance of the RO is likely cheaper.
Test Kits Reagents
You will need replace reagents for the test kits. Also, the minimum given above may not be adequate. The typical SW combo kits are not of low enough range for reef work. They will only be of use during the first few weeks of cycling/curing. That estimate assumes that you will acquire the better tests over time or have access to someone else's expensive tests should you need to diagnose a problem.
You will need it to run the pumps and lights. It won't be insignificant. Electric costs vary. Check the KW cost on your electric bill. Add up wattage of all the equipment you are using, pumps 24hrs/day, lights 12hrs/day. Calculate what the electricity will cost. Don't forget cooling, in many areas, you will need either a chiller or will have to air-condition the room where the reef is kept. The lights will generate heat. At a minimum, your AC bill will also go up accordingly. Electricity cost for a mini-reef system could easily be a couple hundred bucks a year.
In some areas, water is expensive. RO units waste several times what they produce in water. This could add a little more expense.
You may want to do water changes in which case you will eventually need more salt. Salt mixes run $10-$25 per 50 gallon mix.
Fluorescent tubes and MH bulbs wear out. Fluorescent tubes are usually okay for nine months to a year before spectrum shifts and output are reduced significantly. Some tubes, like actinics, may need replacement as frequently as every six months. Replacement MH bulbs is recommended about every one to two years (depending upon spectral shift and output degradation). Add up the cost of your tubes and figure in the replacement cost based on the estimated lifetime.
This can really vary. You probably shouldn't have more than a couple fish in the micro-reef and not more than a handful in the mini-reef. The typical fish suitable for a reef will be from $10(small goby or blenny) to $30(small angel or tang). You could spend $300 on one purple tang though. Pieces of coral, decorative rocks, giant clams and other sessile inverts start at around $20 a piece and go to many hundreds a piece. Snails range from about $1/each to about $8/each and are recommended for controlling algae. Other motive inverts like shrimp range from about $10 to $30.

You probably should start with the snails as soon as the live rock is in the tank. You don't have to have any fish if you don't want any. You don't have to have inverts either although that is probably why you set up a reef tank. Just quality live-rock is very of nice to look at but sooner or later you will likely want something else in your tank. The invert stocking will be very incremental and should be. It is not heathly to add a lot of stock at once. You can spread your stocking over several years. You could spend anywhere from say $100 to $750 on the micro-reef and $200 to $10,000 on the mini-reef.

The official home site of the FAQ is maintained by Erik Olson as part of the The Krib.