Safeguarding Your System from Meltdown -
Stopping Murphy in His Tracks: Part II

Equipment and System Design Related Problems

I should mention at the beginning of this second article that I have a belief about reef tanks: except for an occasional isolated infection that might take out a random coral, reef tanks do not "crash" without a reason. I've read quite a few accounts of reef tank disasters that start with statements like, "My sand bed crashed and killed everything," or, "All my Acropora just died/sloughed for no reason," or, "Everybody knows that Xenia just crash." When I see a post like this I ask a series of questions. In most cases after just a few questions I can get a pretty good idea of what went wrong. Provided there are no instant, obvious reasons (tank too hot, loss of circulation, over-dosed some chemical) I'd estimate that 80 percent of the time corals start dying, it is because alkalinity has dropped too low, and the hobbyist is slow to correct it. Low calcium and, sometimes, low magnesium levels also cause problems, but less commonly, in my experience. While it might not be considered a 'crash' because it might happen over days or weeks, a cascade effect can occur in which the death of one coral pollutes the water and continues the spiral of water quality degradation. I've kept Xenia for about nine years now and raised hundreds of colonies, in a half-dozen tanks that were never consistently supplemented with any form of iodine, and I've never seen a colony suddenly "crash" without an obvious reason. I have, however, seen colonies fare poorly when introduced into certain tanks, possibly due to the presence of toxins from other corals, low lighting conditions or other environmental problems.

Hardware/Equipment Related Problems:

Temperature Control

Both low and high temperatures can injure tropical marine organisms (Shimek, 1997). While the debate regarding the best temperature for maintaining a reef tank likely will never end, most experienced hobbyists become very nervous when their water temperature exceeds 84° F. Designing a cooling system for a reef tank that minimizes the risk of overheating takes careful planning. Lights should be suspended as high as possible, balancing the reduction in the amount of light reaching the tank's inhabitants against the heat they transfer to the water. Allowing heat to escape freely from the top of the lighting system can also help as well as designing luminary systems containing proper ventilation. Fans that blow across the water's surface can have a significant evaporative cooling effect, but this must be weighed against the increased noise and ambient humidity they generate. Chillers may be necessary, along with extra fans placed above the sump. Whenever possible, more than one piece of cooling equipment should be used, and high temperature alarms should be engineered into the system. High temperature alarms may be simple, inexpensive temperature probes with small beepers, or sophisticated temperature sensing and control systems that can shut down lights, sound alarms and even call the hobbyist on a cell phone, relaying the tank's exact temperature. As with other safeguarding systems (see the first article in this series), the choice of how much to spend on monitoring equipment should be balanced against how much could ultimately be lost due to the occurrence of a severe overheating event. As a final point on cooling systems, when using a remote chiller, remember that the reliability of the cooling loop as a whole depends on the pump's reliability in running water to and from the chiller.

The lighting system used on the author's 400-gallon tank is completely open at the top, allowing excess heat to escape into the room. No fans are used near the tank, in order to minimize room noise.

The canopy of the author's tank utilizes guillotine-style removable openings which, when removed, allow complete access to the tank and which, when in place, prevent jumping fish from escaping. These openings also keep the lights' glare from interfering with the view of the display.

Failure of heating control systems can also result in the loss of tank occupants. This is most likely to be caused by a malfunctioning thermostat in the heater causing the heater to remain on, or by a complete failure and breakage of equipment. Submersible glass heaters are reasonably safe in most tanks if they are kept away from anything that might break their outer glass shell. The tank can be contaminated with copper or other toxic compounds from some heaters if this shell breaks, or if its seals fail. Also, over time, most plastics exposed to saltwater will become brittle, and care should be taken that the cords of old heaters are not overly stressed by repeated bending. The best way to minimize overheating risks from heaters is to first adjust the heater's temperature set point to the proper range, and then plug the heater into a separate temperature controller. Even if either device's thermostat fails, an over-heating event still should not occur. When the breakage risks of glass heaters are deemed too large for comfort, consider the recently available titanium, or even plastic, encased heaters. Plastic encased heaters, however, are too new to the hobby to assess their reliability. The titanium heaters are more expensive (again, weigh the risks against the value of your tank's inhabitants) and unfortunately, in my experience and that of a member of my local reefkeeping club, the reliability of hobbyist grade titanium heaters' thermostats appears, thus far, to be only fair.

This system-monitoring device has inputs for temperature (tank and/or room), leak detectors, high- or low-level detectors and other alternate detectors. It can also detect AC power loss and high sound levels (consequently, it could respond to other devices that sound alarms). The system can call out to a list of phone numbers (cell phones, pagers, etc.) and describe the problem that is occurring with a synthesized voice. Alarms can be disabled remotely over the phone, and you can "listen in" to your system with the microphone in the base unit.

Loose Powerheads

While the suction cups supplied with an inexpensive powerhead might work fine for a few weeks, they should never be relied upon as a long-term mounting solution in a reef tank. Most flexible plastics submerged in saltwater become brittle and inflexible over time. A powerhead falling into a tank and stirring up a deep sand bed could potentially cause a chain reaction in a reef tank, resulting in the loss of sensitive corals from a tissue sloughing event (to be discussed in Part III of this series). To prevent this from occurring, powerheads should be secured to the tank's top edge using their supplied plastic brackets or by means of the newly available magnetic powerhead holders. If a powerhead needs to be positioned lower in the tank, a custom-made bracket can be fabricated from a thin strip of acrylic, a propane torch to warm and bend the acrylic and some nylon nuts and bolts to attach the bracket supplied with the powerhead to the custom-made hanger. While perhaps not the safest suggestion, when using the suction cups supplied with the powerhead, simply affixing the power cord to something outside the tank can take some of the weight off the suction cups and help prevent a powerhead from falling.

Overflowing Skimmers

Malfunctioning skimmers often overflow, resulting in a large volume of water in the waste container. In systems relying on a main pump for circulation, this could result in a significant loss of water volume from the system, potentially stopping the tank's circulation as the main pump runs dry. Auto-shutoff waste containers can restrict the skimmer's outflow when the waste container is full, but even these devices are not foolproof, and can sometimes leak. I have begun putting my skimmate waste container into my sump. In the worst case, if the skimmer overflows and the auto-shutoff fails, some skimmate will flow back into the tank. Skimmate flowing back into the tank is certainly not a good thing, but the alternative, losing the tank's circulation, and consequently some livestock, seems much worse to me.

Calcium Reactors

Many aquarists now use calcium reactors as a method for replacing the calcium and carbonate taken up by corals in the process of calcification. The most common thing to go wrong with a calcium reactor is the plugging of its outflow line. This is particularly likely to happen if the calcium reactor is not fed by a high pressure pump, and therefore is not under some positive pressure. The outflow often becomes plugged when a needle valve is used on the effluent line. I've found that it is sometimes better to control a calcium reactor's outflow with a variable length of small diameter tubing than with a needle valve. The longer the tubing, the more it restricts the flow. On a tank with a lot of corals, and that consequently has a high calcium demand, the plugging of a calcium reactor can lead to problems with low alkalinity in as little as a day or two.

The carbon dioxide (CO2) bottle's regulator can also be a source of problems. Carbon dioxide is a liquid at room temperature and under pressure. This means that the level gauge on a compressed cylinder of CO2 will read the same number continuously until about 99% of its contents have been used. Then as it's depleted, in a very short time the cylinder will be empty. Lifting and sloshing the cylinder back and forth can help to estimate the amount of CO2 left. I've noticed that when a CO2 bottle is just about completely empty, the regulator has trouble controlling the delivery pressure. Counterintuitively, the delivered gas' pressure tends to increase just as the tank is about to become completely empty (this is sometimes how I notice my CO2 tank is getting low). As the delivery pressure increases, the flow rate of CO2 to the reactor increases, increasing the calcium and alkalinity output. I notice the outcome as an increase in algae growth on the glass over a few days, possibly due to the increase in nutrients that comes from the substrate being dissolved faster or from more dissolved CO2 in the tank's water. Since normally the line that delivers CO2 to the calcium reactor is a low pressure line (5 psi or so), a hobbyist often might not secure it well. When the pressure unexpectedly increases as the cylinder empties, the delivery line can blow off. If this happens, depending upon the design of the calcium reactor and the way it was plumbed into the system, water under pressure from the calcium reactor could come rushing back out the CO2 delivery line (flood potentials!). Also, the remaining contents of the CO2 bottle might be quickly released. If the CO2 is released in a small enclosed area it could suppress the pH of the tank, or merely cause a dangerous situation by displacing much of the air in the room as CO2 is heavier than air.

Metal Halide Bulbs

Some metal halide (MH) bulbs emit high levels of ultraviolet (UV) radiation. Normally, most of this radiation is filtered out by the outer glass envelope of a mogul style MH bulb or, in the case of a double-ended bulb, by the tempered glass in the fixture. If these filtering glass pieces are absent for some reason (e.g., by a fish splashing water onto a hot bulb or glass shield and cracking it), large scale "burning" of the tank's inhabitants by the UV radiation can occur in a very short time span (from hours to possibly minutes). Be sure to check your metal halide fixtures on a regular basis, as it can take many months for some corals to recover from an exposure to high levels of UV radiation.

Ground Fault Interrupters (GFIs)

While these are an important safety device capable of saving people from a life-threatening electrical shock, hobbyists should be aware that these devices occasionally trip without good reason. I have found the type of GFI that can be purchased as an extension cord to be particularly problematic.

System Design Related Problems:

Overflowing Sumps

This problem can occur during a power outage if a siphon forms in the lines normally returning water from the main pump to the tank. These lines should not discharge their water too far under the surface of the tank's water or, alternatively, small siphon break holes should be drilled in the return line just below the water line. A small amount of water will continuously discharge from these holes, but if the pump stops, air will quickly rush into the line once the tank's water level drops, preventing a siphon from forming. These small holes must, of course, be kept free of debris, and their "siphon preventing ability" should be checked periodically. Remember also that a siphon could form in nearly ANY piece of tubing, however small, running into a tank and back out at a lower elevation. It's a good idea occasionally to pull the plug on all the electrical devices in the tank and check for leaks and overflowing sumps while all the systems drain down to their equilibrium point. This is also a good time to check to make sure that all important pieces of backup equipment power up successfully after being off for several minutes. Of course, monitor the organisms to make sure they do not suffer in any way during your testing.

The pressure activated level sensors mentioned in the first part of this series can be configured to sense either high or low water levels. When I recently replaced some old AC smoke detectors in my home I found that I could rewire them to emit the alarm signal whenever they are plugged in. The combination of these alarms with the level detectors and, ideally, with the high-level sound detection capability of the previously described tank monitoring system, can trigger a phone call alert if the sump's water level drifts too high or low, for whatever reason. Simple and inexpensive conductivity-based leak detection alarms also can be positioned in a sump to sense a high water level.

Screening Intakes

All large or vitally important main circulation pumps should have screened intakes. Any foreign object sucked into a pump has the potential to drastically decrease the water flow, or even stop it completely. A hard object sucked into a pump can even break a fin off the pump's impeller, causing the pump to become unbalanced and potentially, to shut down. Also remember to safeguard overflows from fish and motile invertebrates such as anemones or snails. To protect these organisms it is best to try to spread the water's intake over an area as wide as possible, thereby minimizing the flow rate over any one area.

Gate Valves on Overflows

In an attempt to control the flow rate through their systems and to control water noise, some aquarists have take to installing gate valves on the lines exiting their overflows. Gate valves are a bright idea, but be careful. Using a gate valve on a line that's downstream of an overflow can render the overflow perfectly silent, as no air is entrained into the water flowing to the sump. The water's height in the overflow, and therefore the water's driving force or head, can naturally compensate to some extent for minor flow restrictions in the line. If a gate valve is used on an overflow line to a sump there must be an alternate pathway for the water to flow to the sump when, and if, the gate valve clogs, or else a flood is inevitable.

Salt Creep - High Humidity

To reduce the buildup of salt creep, try to minimize the number of pieces of equipment which contact the water's surface, extend out of the tank and contact the aquarium's edge. During periods of very high room humidity in the summer, be particularly wary of salt creep near electrical outlets and sensitive electrical equipment. Deposits of salt can attract moisture from the air and form a dangerous conductive condition, as well as a corrosive substance. For this same reason, try to minimize any splashing of water as it enters a sump.

Localized Heat from Metal Halide Bulbs

Be aware that the localized mechanical stresses from multiple heating and cooling cycles can take a toll on materials located near metal halide bulbs. Acrylic tanks can form cracks that can propagate across and even down the face of a tank. Glass tanks also can have problems. In recent years tank manufacturers have begun using thinner glass. They compensate for this thin glass by adding strips of plastic across the tank's top frame. These braces are likely to work fine for most tanks sold that use low wattage fluorescent lights, but they often do not stand up to high intensity reef tank lighting. Raising the bulbs or moving them away from these braces is the simplest solution. Gluing a sacrificial piece of thick acrylic over the brace, or replacing the brace with some stronger structural support, are also practical options.

The next and final article in this series will cover unpredictable events, preparations for an extended absence and some final miscellaneous tips I've learned from my numerous brushes with disaster.

If you have any questions about this article, please visit my author forum on Reef Central.


Shimek, R.L. 1997. What are Natural Reef Salinities and Temperatures… Really… and Does It Matter? Aquarium Frontiers.

Reefkeeping Magazine™ Reef Central, LLC-Copyright © 2008

Safeguarding Your System from Meltdown - Stopping Murphy in His Tracks: Part II by Greg Hiller -