DIY Universal Battery Backup For Your Reef

[WWIII]

Great write up @chipmunkofdoom2!!! If I can't convince the wife we need a whole home generator, I will be following this guide to install a battery backup very soon! Thanks for taking the time to do this!

 

[chipmunkofdoom2]

Great write up @chipmunkofdoom2!!! If I can't convince the wife we need a whole home generator, I will be following this guide to install a battery backup very soon! Thanks for taking the time to do this!

No trouble at all, I'm glad it was useful to you! Feel free to pop back in if you have any more questions

 

[nero0762]

Great writeup!
I'd like to get some more input/thoughts on the bare minimum equipment needed. I know the answer depends on the length of the outage, so let's assume 24-48 hours...
Originally it was suggested that a return pump, powerhead(s), and possibly a skimmer, and maybe even media reactors.
I'm wondering if you really need any more than your powerheads, and a source of oxygen (air pump, etc)
I get that without the return pump you give up your mechanical and chemical filtration. But you'd still have the majority of your biological filtration, right? So the water may get a little "dirty" to the eye, but unless you're running a very heavy fish load in the tank, I'd think it would survive just fine for 24-48 hours, wouldn't it?
And, I would think by reducing the equipment down to a few powerheads and an air pump, the capacity needed would be significantly smaller (or your run-time would be significantly longer).

 

[chipmunkofdoom2]

Thanks @nero0762. You're correct on all accounts. The goal should be to minimize equipment running. That will make your batteries last the longest. In an outage, the first priority is to get water moving. This keeps it oxygenated and keeps areas of the tank from becoming stagnant. After the water's moving, you can think about more advanced things, like keeping water flowing through your sump, running a skimmer, and potentially running reactors. You'd only want to consider running media reactors or a skimmer if you absolutely needed the filtration. As you say, that probably wouldn't be for about 2 days or more. In that case though, you should be looking for a generator anyway, because heat will start to become a problem after an outage of a few days.

In my case, I run a Jebao PP-4 in my display and a Koralia Nano 425 in my sump to keep things going down there as well. That's about 11W total (7W for the Jebao, 4W for the Koralia). On a 50Ah battery, if we assume an 80% depth of discharge and add 10% to account for inefficiency, I'd get about 40 hours of run time. If I were to add my return pump, which I believe uses 14W, that would drop my run time down to only 17 hours. If we were to add my skimmer too, the run time would drop even further to about 11 hours. If I had a 100Ah battery, that would effectively double the run times (80 hours with just the circulation pumps, 34 hours if you add the return pump, 22 hours if you add the skimmer too).

You can basically run anything you want to if you buy enough battery capacity. You could even wire this system into your circuit so your entire tank keeps running in the event of an outage (@Oleshp discusses doing this on the first page). While this wouldn't be my personal choice, there's nothing wrong with doing so if have the budget and if keeping everything running is really important to you. You can also get a lot of bang for your buck if you're smart with your planning. A 50Ah battery costs about $100. So for that, plus a $30 inverter, plus a $15 charger, plus a cheap DIY transfer switch (or the more expensive $60 commercial option) you could easily run two smaller pumps like the one I describe above for 40 hours. Not bad when you consider that some people spend that much on a tiny frag of coral.

 

[Trueblue17]

You're welcome

Correct, the inverter would need to be always switched on. I've wanted to put a relay between the battery and the inverter, and only have power flow when the utility fails. But, I haven't thought of anything good yet. The problem is that because the battery supplies 12VDC, the current flow can get pretty high from the battery to the inverter, even at relatively low wattages. I haven't found a relay with a high enough capacity to put between the battery and the inverter. Not one that I would feel comfortable enough with.

To make sure the inverter is working, I do a system test once a month or so. I unplug the utility cable and make sure the switch goes to battery backup and that the inverter is working properly.

ya that would be cool to figure out a way to not have to leave the inverter on, I have 2 100watt solar panels running to ,2 golf cart batteries connected toa 2000watt inverter that goes to an automatic transfer switch ,to a gfci outlet I made up lol works great , just the inverter still drains batteries slowly if inverter is left on all the time

 

[Daltrey]

I've been using a deep cycle battery backup for over a year now with no problems.

 

[Saltyanimals]

Appreciate this writeup. This has been on my to do list for awhile and was looking for a updated detailed write up. Thank you

 

[chipmunkofdoom2]

ya that would be cool to figure out a way to not have to leave the inverter on, I have 2 100watt solar panels running to ,2 golf cart batteries connected toa 2000watt inverter that goes to an automatic transfer switch ,to a gfci outlet I made up lol works great , just the inverter still drains batteries slowly if inverter is left on all the time

How much of a draw does your inverter put on the batteries? I know it's greater than zero, but is it that much? If you have solar panels hooked up, I would think they would be sufficient to keep the batteries topped up. The reason this is so tricky is because of how much current the inverter must draw to make AC power. 500W is only 4A at 120VAC, but it's a staggering 41A at 12VDC. If you were to shut off the inverter with a relay, it would have to have a really high capacity.

I did think up a rather convoluted way of shutting off the inverter and only turning it on in the event of an outage, but I've never built one or tested it. You would need a 120VAC SPDT relay and a high-current automotive starter relay. You would wire the 120VAC relay to your utility power. Connect the positive of the battery to the NC (normally closed) terminal of the 120VAC relay. Then, connect the Common terminal of the 120VAC relay to the positive of the starter relay. Run a negative line from the battery to the negative of the starter relay. Then, run the positive from the battery through the starter relay and to your inverter. Finish up by wiring your inverter to the negative of your battery. Here's a sketch of the system:

When the utility is supplying power, the 120VAC relay is open. The NC terminal is disconnected, which means power does not flow to the starter relay. Because the starter relay sits between the inverter and the battery, this means no current flows to the inverter. When the utility power fails, the 120VAC relay closes. This connects the battery to the starter relay, which opens the starter relay, which connects the battery to the inverter.

There are a few problems with this type of a setup. First, I don't know if automotive starter relays are rated for continuous operation. Typically, a starter relay is designed to supply a very high current to the starter motor for a few seconds. I don't know if continuous operation would wear it out prematurely, or worse, cause it to fail catastrophically. Second, and probably the bigger issue, is what happens if the relays fail. If either of these relays fails, your inverter may not get power in the event of an outage, which is a pretty big deal. This is a safety net, so it has to work when the time comes.

For these reasons, I chose to just leave my inverter on 24/7. It's hooked up to a battery maintainer, so the battery is always floating at 13.2VDC anyway. If the draw from the inverter at idle is truly that significant, you may want to consult an electrician and ask them how to do this properly. There's likely something an electrician knows that I don't.

 

[chipmunkofdoom2]

I've been using a deep cycle battery backup for over a year now with no problems.

Looks good! My one battery is an Exide also it's lasted me a good 6 years or so and saved my bacon during the last outage.

Do you have your DC pumps hooked directly up to the battery via their battery backup ports?

 

[truepercs]

Lithium batteries certainly are superior in many ways to lead acids. I definitely think lead acid batteries are a better choice for aquarium emergency backup though, for several reasons.

First, you're only going to be drawing from the batteries when your power is out. Unless you have extraordinarily unreliable power, this is likely only going to be a few times a year. Lithiums might give you a few thousand charge cycles as opposed to the 100 - 200 of a lead acid, but if you only discharge the system two or three times a year, in 10 years you'll only burn through 30 charge cycles. Paying more for charge cycles that you'll never use doesn't make financial sense to me.

Second, size and weight are definitely an advantage of lithium batteries. But, again, most reefers are likely setting up a battery backup in a closet and leaving it alone. It doesn't need to be light or especially compact, nor does it need to fit in a drawer (although that is nice if space is tight).

Third, accessories for lead acid batteries (lead acid chargers, 12V inverters, etc) are all over the place and are reasonably cheap. The stuff for lithium batteries is harder to come by, and is usually a bit more expensive.

If you have an application where you're going to be charging and discharging your batteries a lot, lithiums are critical. The same is true where space or weight is an issue, like with Tesla's Power Wall or electric cars. If you already had some Li-ion batteries, that's great, and they would certainly work for this purpose (I am truly envious.. the most I've ever done is played around with some 18650 cells. I'd love to get my hands on some larger-capacity Li-ion batteries). However, when you're going for cheap and the system doesn't have to be deep cycled that often, I still think that lead acids (AGMs in particular) are the way to go.

Lithium batteries certainly are superior in many ways to lead acids. I definitely think lead acid batteries are a better choice for aquarium emergency backup though, for several reasons.

First, you're only going to be drawing from the batteries when your power is out. Unless you have extraordinarily unreliable power, this is likely only going to be a few times a year. Lithiums might give you a few thousand charge cycles as opposed to the 100 - 200 of a lead acid, but if you only discharge the system two or three times a year, in 10 years you'll only burn through 30 charge cycles. Paying more for charge cycles that you'll never use doesn't make financial sense to me.

Second, size and weight are definitely an advantage of lithium batteries. But, again, most reefers are likely setting up a battery backup in a closet and leaving it alone. It doesn't need to be light or especially compact, nor does it need to fit in a drawer (although that is nice if space is tight).

Third, accessories for lead acid batteries (lead acid chargers, 12V inverters, etc) are all over the place and are reasonably cheap. The stuff for lithium batteries is harder to come by, and is usually a bit more expensive.

If you have an application where you're going to be charging and discharging your batteries a lot, lithiums are critical. The same is true where space or weight is an issue, like with Tesla's Power Wall or electric cars. If you already had some Li-ion batteries, that's great, and they would certainly work for this purpose (I am truly envious.. the most I've ever done is played around with some 18650 cells. I'd love to get my hands on some larger-capacity Li-ion batteries). However, when you're going for cheap and the system doesn't have to be deep cycled that often, I still think that lead acids (AGMs in particular) are the way to go.

It is more of an experiment in using solar and batteries to run a system...

 

[cvicente]

A lot easier to use a 1500VA UPS with an external battery bank. My set up gives me 8 hours on the main pump and 7 days on the Vortechs

 

[chipmunkofdoom2]

A lot easier to use a 1500VA UPS with an external battery bank. My set up gives me 8 hours on the main pump and 7 days on the Vortechs

I'm not aware of any UPSs that allow for external battery sources. Likely, you'd need to disassemble a working UPS to hook up external batteries. Even if there are UPSs that accept external batteries, they'd likely be very expensive. Plus, that basically makes the device an inverter/charger. I'm not a huge fan of consolidating so much of the backup system into one device, whether it's a UPS or an inverter/charger. I much prefer making my own system with interchangeable parts that can be swapped out for any other off-the-shelf part with similar functionality and capacity.

Easier is subjective. What's "easier" will vary from person to person. For me, ordering cheap, commodity parts off Amazon and hooking them together was easier than buying an expensive (or used) UPS and disassembling it. A UPS with external batteries would certainly work. If you already have a quality UPS, this might be a good option. It's just not my favorite for the reasons I've outlined previously.

 

[gcrawford]

I'm not aware of any UPSs that allow for external battery sources. Likely, you'd need to disassemble a working UPS to hook up external batteries. Even if there are UPSs that accept external batteries, they'd likely be very expensive. Plus, that basically makes the device an inverter/charger. I'm not a huge fan of consolidating so much of the backup system into one device, whether it's a UPS or an inverter/charger. I much prefer making my own system with interchangeable parts that can be swapped out for any other off-the-shelf part with similar functionality and capacity.

Easier is subjective. What's "easier" will vary from person to person. For me, ordering cheap, commodity parts off Amazon and hooking them together was easier than buying an expensive (or used) UPS and disassembling it. A UPS with external batteries would certainly work. If you already have a quality UPS, this might be a good option. It's just not my favorite for the reasons I've outlined previously.

I've used a modified UPS (like the one in the previous pic 1500 I believe) and connected them to two sealed lead acid batteries with much success. It would worry me to have an inverter running and laying on my floor 24/7. With a large UPS, the inverter activates automatically. Also, my UPS filters the incoming AC voltage and protects against surges which is not a huge benefit since I only have my vortechs connected to it, oh and my APEX 12v backup power supply.

Neat ideas though. I would be very cautious about using Lithium batteries.

 

[Daltrey]

Looks good! My one battery is an Exide also it's lasted me a good 6 years or so and saved my bacon during the last outage.

Do you have your DC pumps hooked directly up to the battery via their battery backup ports?

Yes, It works awesome. As soon as power is lost it automatically switches to battery backup. It also shows that it's running on battery backup thru the Apex app.

I do alot of kayak fishing and use the battery for my kayak trolling motor also.

 

[Cresta]

chipmunkofdoom2,

Wow, what a great thread and detailed write-up.

I build something very similar to your setup back in 2002.
http://www.reefcentral.com/forums/showthread.php?t=64918

I didn't start that thread. I was one of the guys that was asking questions. I build one also with a relay as the transfer switch. I just used a regular wet cell car battery. I researched and found float charging the battery wouldn't give off much harmful gas, so I didn't purchase a much more expensive sealed battery. I did enclose the battery in a plastic case with tie straps to contain the battery if it was to explode. I remember I bought it at Pep Boys....and it was design for that purpose of storing battery. I have seen battery explode under the hood of a car and its very scary. The plastic case will contain all the acid spill from the battery if it was ever to happen.

I used my backup creation for about two years and took it apart to upgrade and never finished the project again..... forgot about it until reading your thread now. Back then, I didn't calculate the current drawn by my equipment and the backup was only able to sustain my return pump for about 30 minutes. I wanted to increase backup time by adding another battery and also adding a second relay. But....like I said...never got around to finish it.....These days I'm using DC pumps and building backup power source is a lot easier to do.

Anyways, thinking back....I also used a OMRON relay as a transfer switch. I initially had it mounted on a relay base. So if it needed replacement, I pulled off the old one and replace with new, as the wiring are connected to the base. I wouldn't need to unplug each spade connector to reconnect a new relay. After a while I realize, if the relay fails... the whole system fails. Even monthly testing wasn't enough, as I had tested the setup 2 days prior to an outage and then system didn't kick in when needed. The heart of the system relies on the $2 relay. For redundancy, I wanted to connect 2 relays to the system. That way even if one fails, the second will still complete the circuit. Do you have any suggestions on the wiring of dual relays? I remember that was what I got confused on and didn't finish my upgrade.

You mentioned the commercial unit Xantrax for $65 and the GO Power 30A transfer switches..... what is the main difference between the two? 30A vs a smaller sized switch on the Xantrax?

Your insight is highly appreciated,
Ermin

 

[salty joe]

Very well written plan, very informative, thank you!

 

[chipmunkofdoom2]

Thanks for the compliment @Cresta

You bring up some really good points with the DIY transfer switch. I used insulated disconnects to connect my wires to the relay. If the relay did die, I could simply unhook it and hook a new one up. But, a base into which you could plug and unplug the relay would be much simpler. I'm actually thinking up a new system right now that's much more plug-and-play, perhaps I'll get a base for the relay to make wiring it a bit easier.

You mention that your relay failed. Did it fail with NO connected to common? How much load were you running through the relay? It's my understanding that if an elecromechanical relay fails open, it's because the contacts fused upon opening. It's my understanding that this really only happens when you try to run a lot of current through the NO terminals.

Dual relays are an option, for sure. In a one-relay setup, you wire the Common to one wire (or outlet). In a dual-relay setup, you would wire both of the Commons to one wire (or outlet). The NO and the NC would be wired as usual. My concern with dual relays is what would happen if one was open and the other failed closed, or vice versa. If a mechanical double throw relay fails, it never fails "off," in a state where no power gets through: at the very least, one connection is always made (NO to Common or NC to Common). Let's assume the coil failed in one of the relays: that would mean the coil wouldn't energize and NC would be connected to Common (the inverter would be connected to the load, in other words). The other relay, if it continued to operate as it normally would, would be supplying utility power to the load. The net effect is that you would be essentially connecting your inverter to the utility. I don't know what unexpected consequences this might have. It's possible this could damage your inverter, or even cause a fire. I'm not an electrician and that's a little outside the scope of my understanding.

There are definitely cons to having one relay. If it fails, your system doesn't work. I'm just not sure how the system would behave if both relays were operating in different states, which is likely to be the case if one fails. Perhaps an electrician can comment? If you wanted to use two relays, perhaps wire them to separate outputs. That way, if one relay fails, the equipment hooked to that relay just won't run. The other relay will continue to operate the equipment connected to it. I haven't thought of a good way around this problem, but I'll try to think of one.

As far as the Xantrax vs the Go Power! transfer switch, I think the only difference is really the current. The poster was concerned about hooking up a 15A switch (the Xantrax) to their circuit. The circuit has a maximum draw of 20A. The circuit would see no problem providing 20A to the Xantrax, but the Xantrax would fail under that much power. If, however, the Go Power! switch was used instead, the circuit would hopefully trip if it tried to provide more than 20A to the switch. Because the Go Power! is rated for up to 30A, the circuit would trip before the switch was overloaded.

 

[Cthehentz]

Is the inverter a modified or pure signwave?

 

[chipmunkofdoom2]

Hey @Cthehentz, I'm using a modified sine wave inverter. Pure sine wave would work, but in my opinion it's not necessary in the context of short term battery backups. My thoughts from a previous post:

The only example I can think of where a pump would not run on MSW (modified sine wave) is Rossmont's controllable AC pumps. According to company reps, the pumps are controlled by altering the sine wave coming from the AC power source before feeding it to the pump. A dirty sine wave might not perform as well, if at all. Other than that, almost all aquarium pumps should run without pure sine wave (PSW). PSW is undoubtedly better, but I don't think it's necessary in the context of aquarium backups.

The reason MSW (modified sine wave) is harder on AC motors is because it's not a smooth waveform. The motor's natural movement relies on a smooth up and down sine wave. When you feed an AC motor a jerky and sharp modified sine wave, it's not going to run as smoothly and will produce extra heat. The heat is not a huge concern in our applications, as the motors are immersed in a liquid coolant (our aquarium water). That's not even considering that a lot of pumps today are powered by an AC/DC power supply. The only effect the MSW might have on DC pumps is the AC/DC supply runs a bit warmer.

Over time, the lack of a smooth sine wave could shorten the life of AC pumps. But remember, this system is only going to be running the pumps during a power outage. If you're only running your pumps on this inverter for 10 - 20 hours out of the entire year, how critical is it that the pump be run on pure sine waves? If you have a lot of outages and very expensive AC pumps in your system, a PSW inverter might be worth it. I personally feel very comfortable using a MSW inverter.

 

[SteadyC]

Nice post! I did almost all of this about two years ago, everything minus the switch. I do have a computer UPS for automatic short term coverage. Then I can manually switch to my system, picture here. Battery tender (left) and inverter (right) on the second shelf, battery on top. Mine is a sealed dry cell lead battery, 5000 amp. I have a power strip in fish room, with an extension cord that is ran to where this all sits. I plug the extension cord in the inverter, connect the inverter to the battery, then move the pump and heater power cords to the power strip. Tank power heads are on their own UPS’s.