System Reliability

Stanley Brown

April 15th, 1999 on #reefs

Stanley Brown was first introduced to aquariums in 1961 when he was 9 years old. His first "saltwater" aquarium was in 1971, then ran a 1,500 gallon wholesale business from 1978 to 1980, after which he joined California Fish and Game, Inland Fisheries division and now works in the water purification/process industry. He has written articles for FAMA, MFM and SeaScope, spoken at MACNA and WMC conferences, and the 8th International Coral Reef Symposium (Panama). Stan is currently the President, CEO of The Breeder's Registry, and the editor of the Journal of MaquaCulture. He's current main interest is propogation and larval development (i.e. making numerous attempts to spawn marine organisms with some success).

How well a system "works" is more than just luck or having a "wet thumb". My marine aquarium experience now spans 28 years and I have owned and witnessed many systems using almost every "latest and greatest" methodology. Some have proven more successful than others, however in retrospect, in some instances the poor performance of some applications may have been due more to design flaws than a less than satisfactory methodology. I have been paying closer attention to why aquarist experience loss of livestock since the conception of the Breeder's Registry (now in its 7th year). The majority of problems associated with the loss of animals which had begun to spawn, or corals which were exhibiting sufficient captive growth to fragment or make cuttings is not disease, but a failure in the mechanics of the system.

We rely on devices to maintain and control flows, photo period, monitor and maintain water parameters which together is essentially a `life support system'. These devices require an energy source and many have moving parts. Some can be "serviced" while others cannot. ALL have a finite life. Those that can be serviced often are not and typically the maintenance practices amount to "If it ain't broke, don't fix it!" and "When it does breaks, repair or replace it. (usually the latter)" The latter event is often the aquarists downfall, resulting in loss of precious livestock which equates to a loss of time, effort, and money.

In an effort to have the "best" system we all too often end up with a system which is a series of modifications and additions. This may result in a system providing an excellent environment for the animals we choose to maintain, but the ability to easily service the equipment may be another matter altogether. We assume that things will work for ever so little thought is given to how we will service, repair, or replace a malfunctioning or failed component. Servicing can be something as simple as cleaning, both internally and externally, or more complex involving the replacement of seals, o-rings, and other parts. If you are in the process of building a new system, some additional planning "up-front" may make your life much easier "down the road". If your system is already in operation, developing a maintenance plan with a contingency for component failure could minimize your risk of losses.

Regardless of what methodology you ascribe to, almost all systems;

• Require electricity
• Will use one or more pumps to move water
• Have a heating and or cooling device
• Have some type of lighting

Electrical requirements are often overlooked and under-estimated. Power outages by the utility company is out of our control but circuit overload is within our control. IF there is an available place to plug something in, there is "plenty" of power. Quite often a single electrical circuit is supplying power to operate the pump(s), heater/chiller, lights, and whatever else may require electricity to operate. In addition, the electrical outlet for the aquarium is one of several outlets on the circuit. This circuit will have a load protection in the form of a fuse or a circuit breaker and is generally rated at 15 amps and occasionally 20 amps. The "load" is the amount of power used by the devices connected to the circuit (plugged in). If the load exceeds the rated protection the fuse will "blow" or the breaker will "trip".

Fuses must be replaced and purchasing additional fuses and keeping in a place (that's easy to remember) is your first step in minimizing risk. Fuses do not get "weak", they either work or they don't. A circuit breaker has the advantage of being reset and power quickly restored. Circuit breakers can become "weak" and trip prematurely (in which case they must be replaced). IF you experience power outages due to frequent circuit breaker or fuse failures you should check each device and connections until the problem is resolved as this could become potentially dangerous (fire hazard). Power strips are often used and typically come equipped with an internal circuit breaker (which doubles as a switch). The disadvantage is if the power strip circuit break should trip, everything plugged into it is without power.

An alternative to having all "your eggs in one powerstri.... I mean basket" is to install inline fuses for each of the devices. The power cord (which is usually "zip cord" ) is "split" and one of the wires is cut. The inline fuse is attached (typically soldered) in between the ends of the cut wire. A properly sized fuse is inserted into the fuse holder. If the piece of equipment has a covering or mounting box, a panel mounted fuse can be used instead of the in-line cord type. Now the individual piece of equipment has load protection without jeopardizing other parts of the system. To select the proper use a fuse which is typically one size above the rated amperage/current draw. Since some (OK many) pieces of equipment are rated in "watts" , divide the wattage by the household voltage (110, 120 VAC, etc) and this will give you a value in amperes (example; a 175 watt MH operating on 120 volts equals an amperage value of 1.458 amperes). In reality, the ballast is not 100% efficient and will also use some power (10% is a good rule of thumb for tar type ballast) so the power load for this equipment would be ~ 1.6 amperes. The fuse needs to sized greater than the load (otherwise it may fail every time the device is turned on) so a rating of about 1.2 times the load is typically used (motors may require an even higher factor). So now we have a value of 1.92 amperes for the fuse rating for the MH. Since this size fuse is uncommon (if even available) a 2.0 amp fuse would be selected. In larger set-ups where you have outlets located for a specific piece of equipment, fused receptacles are available (from the hardware store) which replaces the duplex outlet, having one "plug" and a bayonet base (screw-in) type fuse.

I mentioned that power outages from the utility company are beyond the aquarist control. Some aquarist have opted to add a "battery back-up" (commonly used for computers) or have a gasoline powered generator for these occurrences. This may be an option you wish to look into.

On the topic concerning electrical concerns, there is a very easy and inexpensive means of reducing risk. LABEL your power cords with a corresponding ID on the device. This will save time from having to trace wires and if you should happen to be away and get a telephone call from the "aquarium sitter" it will be much easier to instruct them to disconnect a specific cord. I learned this lesson the hard way (and it took two events to finally have it sink in). I presented a paper in Panama in 1996. My oldest daughter was the "keeper of the fish" while I was gone. I received a message to "Call home ASAP". When I did, she said that something was making a loud noise and thought it was a filter. Needless to say, at dollars per minute for the phone call, having labeled the equipment would have made the resolution much faster, easier, and cheaper! (fortunately it was not a major problem. this time). Some time later, my wife called me at work to inform me the fishroom was filling up with smoke!!! After several "go look at this" exchanges "we" decided it was a MH ballast failing. I told her to unplug it and I would come home. When I arrived home I found that my wife, not being sure exactly which cord was "it" unplugged everything! I promptly labeled all cords, lights, pumps, etc.)

Another maintenance feature which is inexpensive and may allow you to optimize your lighting is a simple hour-meter. The hour meter is wired in series with the timer for the lights and faithfully counts the hours the light is on. I write the hour on each bulb when I install it and I can keep track of usage as well as how long certain bulbs last. Changing bulbs at 6, 9, or 12 months has little to do with the actual number of hours. I try to hold close to 3000 hours, which is ~200 days for SO fluorescents and just over one year for MH. This is a new capability for me so I have no data to indicate whether 3000 hours is a good or bad value. (see Mvc-003s.jpg)

Heaters. how often I have heard the sad tale of fish bouillabaisse or tridacna clam chowder.. The heater didn't turn off and it was too late by the time it was discovered. A simple solution: If your system requires a 200 watt heater, use 2 100 watt heaters. The odds of both heater contacts remaining on above the set point is remote and the smaller wattage heater will not raise the temperature as quickly (or possibly even to dangerous levels).

PLUMBING.. Except for small and self-contained systems, aquariums have some external plumbing; from the tank, to the sump, to the pump, to the. well, you get the idea. If "something" which normally has a water filled piece of tubing, hose, or PVC pipe, starts to leak, or needs to be removed for service or replacement, there is often nothing in the design to allow for the work without emptying or otherwise removing water from the system. Often this involves shutting the circulating sub-system down while the work is carried out.

My "day job" in industrial water process has shown me several ways to operate in a near "zero down time" mode. Plumbing design often seems to be "connect this to that, glue this here, run this over there." I have seen several instances where the pump was installed with all of PVC fittings except for the piece threaded into the pump glued! The pump failed and unfortunately there was no means of removing it without cutting pipe (which happened to be full of water). Had the aquarist installed an isolation valve on the suction and discharge side of the pump, these could have been closed leaving only a small amount of water to contend with when the pump was removed. The use of a "true-union" type ball valve would allow for removal without cutting any pipe. This type of valve is like a union fitting with a valve in between. It can be closed and disconnected at either side. They cost a little more, but I think they are well worth the effort. Additionally, the use of a unions is strongly recommended in any location you think might need to be cleaned out, or possibly modified "some day". BE AWARE that ball valves and unions are not inter-changeable from one manufacturer to the next, so it is best to use one brand. You will need fewer spare valves or unions if you stay with a single manufacturer.

Another issue is pumps. Many systems rely on a single pump for the main circulation. As long as the lump operates, everything is fine. Most local fish stores do not stock spare pump. Many don't even keep the larger pumps in stock and must be ordered. This is not a good situation if you need parts or a replacement pump NOW! Among the possible solutions are:

1. Buy a spare pump (just in case)
2. Buy all the spare parts available (just in case)
3. 1. and plumb the spare pump into the system (parallel) so it is ready to run if needed, allowing you time to deal with the other pump. This assumes that you (or someone) is there when the first pump fails.
4. Instead of one large pump, buy two smaller (half the flow) pumps and install in parallel (as in 3.) and operate both. If one fails you still have partial flow.
5. Go one step farther than 4. and install three pumps in parallel, operate two at a time and the third serves as a "stand-by" ( photograph of this arrangement)

Some may see this as overkill, but the cost of the pumps and valves pales in comparison if the choice is between the pair of fish that you've had for 2 years have finally just started to spawn and the extra dollars.

With the three pump set-up, should a pump fail, I simply valve out the failed pump, valve in the stand-by and plug it in. Then at my "leisure" I contend with the failed pump. This is also something that I could instruct someone to do over the phone.

Another "option" often overlooked is to plumb in a tee'd valve with a garden hose adapter off of the pump discharge line. I have siphoned water from my tanks for years, never even thinking to let the pump do the work! While siphoning is still needed to remove detrital build-up, water changes are typically larger than the amount of "bottom siphoning" required.

A final note is about the "auxiliary" current devices. In my new system I have tried to avoid the use of power heads and wave-makers. The reasons are two-fold; less mechanical devices equals less things to break and eliminate the associated heat transfer. My new system relies almost entirely on a modification of the Carlson Surge Device (CSD). Gravity works in my favor to create surge. The magnetic drive pumps pump everything up to the ceiling and delivers water to CSD's made from 5 gallon square laundry detergent buckets or 20 gallon bottom drilled aquariums which can be isolated from the system and used as quarantine or hospital tanks. I have successfully added a CSD which operates from the overflow of one tank (fed by 2 CSD's), which like all other return lines empties into a 250 gallon sump. Utilizing the random and varying surges in the sump created by the 7 returns the flow is very strong at times and the sump has been fitted as a coral grow-out tank (overview close up. Some people do not like the bubbles created by the CSD, but I don't mind at all and it contributes to the aeration of the system water.

I have tried to learn from my own and others experiences. It is my hope that some of what I have said will be helpful in improving your systems reliability and assuring you success with the animals you maintain in your aquariums.

If I plumb two pumps in parrallel and have them both running and one craps out will the running pump push water back through the dead pump or up to tank?

Yes, and in this configuration you would need to place a check valve on the discharge of each pump.

Does the surge create a lot of bubbles in the water? If so, how can this be remedied?

To me, it is not a lot, salt creep is not a problem. Having the air gap sized correctly will quickly eak the siphon and end it auptly (and less air). How to determine the correct air gap is a trial and error matter, but not to taxing. About air gaps I have found that rather than drilling holes to "suck air" to eak the siphon it works better (for me) to cut the bottom of the siphon tube (in the bucket) square and then cut a 45 degree cut starting on the middle, so half the cut is squared and half angled. Also, IF the discharge is 2" below the surface, you need to build the "bucket" so that 2 - 3 " of water can fill above the top of the siphon tube (in the bucket). ...filling up the airwaves :-)

How do keep up with freshwater replacement due to evaporation in the CSD systems?

Freshwater is presently taken care of with ~ 7 liters of kalkwasser pumped in nightly (mertering pump)

Is there a nice neat inline fuse device you can purchase, instead of having to play with a soldering iron?

All of the in-line fuses I've used have been solder type (but I've mainly got them from Radio Shack (a US electronics chain).

Since you oached the topic of checkvalves is there a reliable one?

I prefer ball-checks over "flapper type". The key seems to be in periodically removing and cleaning in white vinegar or a mild HCl solution. The calcium buildup will cause them to leak by.

Have you attempted to try out any sort of wavemaker type device for powerheads/pumps? What is your opinion of them?

I have used the alternaring timers and they seem to work fine. I've also played with one of the rotating head designs...can't remember the name. It too worked fine. I have tried to get away from powrheads simply from a maintenace perspective (and the fact that when my system is complete it will be ~4000 liter. Too many devices to contend with. I would rather work with the animals :-)

Does the pump arrangment you have on the sump loose much flow due to the way the piping is on the exit side?

There is certainly some friction loss in the plumbing. This was one reason for selecting the MD5s. I've also upsized the piping on the discharge one size to help reduce pressure loss. Having 7 feet of vertical elevation pushes me down on the pump curve! I estimate the two pumps produce about 200 lpm (55 gpm)

Thanks for the great talk, Stan!

© 1999 www.reefs.org