DIY Universal Battery Backup For Your Reef

[kswan]

Thanks for the reply @chipmunkofdoom2. Yes, I'm sure the UPS is 24v. It has 2 12v batteries wired in series. I did purchase the battery just last week. I went with that one because it is the exact same battery that is in the sump pump backup. I was under the impression that this was a maintenance free (sealed) battery. How do I tell?

 

[chipmunkofdoom2]

Thanks for the reply @chipmunkofdoom2. Yes, I'm sure the UPS is 24v. It has 2 12v batteries wired in series. I did purchase the battery just last week. I went with that one because it is the exact same battery that is in the sump pump backup. I was under the impression that this was a maintenance free (sealed) battery. How do I tell?

Hm. Well I suppose I can't say that it's not sealed, although nowhere on the Internet or the battery does it explicitly say it is sealed. The two rectangles on top are the vents that regulate the internal pressure of the battery as it degases, so the battery is at least valve regulated. While that may slow the release of hydrogen, the battery still must release gas if the pressure builds. Additionally, note on the right cap it says "do not tip." This is likely because acid can spill out of the battery is not kept vertical. The warnings about explosive gasses and sulfuric acid are further tipoffs that this is likely not a sealed battery. Both of the AGM batteries I have explicitly say "non-spillable" on the case and state that they can be installed in any orientation. I looked on the O Reilly site and it appears this battery is only around $90. If that's how much you paid for it, then it's not strictly speaking a bad deal. You'll just want to be careful not to discharge it too far, as well as being mindful of the other caveats associated with flooded batteries.

If the APC is 24V, then that complicates things significantly. First, it's unlikely you'd be able to use the battery from the sump backup. That system runs on 12VDC, and by connecting another battery in series to get 24VDC, you'd double the voltage. While I don't have specifics, it would be unlikely that adding a battery to the sump pump would be done in series. This would require lots of additional control electronics. When adding a battery to the sump pump backup, it's likely done in parallel, meaning same voltage (12V), increased capacity. If the electronics in the sump pump battery backup don't fry from being exposed to 24VDC, it at the very least won't work. Second, if you did disassemble the sump pump backup to use the battery in series on the UPS, there's the issue of age and capacity that could affect how long both these batteries last. Third, if you stole the battery from your sump pump, the sump pump wouldn't work during an outage.

If you wanted to use the APC, you could buy another Super Start battery and wire them in series. Then, you could connect them to the APC and it would function normally. It's worth noting that these batteries appear to be 90Ah though, which would be 180Ah total. If that's the case, it's probably going to take the APC a very long time to charge them. They would work though. If you don't want to buy another battery, a DIY system would probably be the next cheapest option. You have a battery, charger, and inverter, so the transfer switch would be the last piece. The Xantrax switch I linked in the main post is $65, but you could DIY cheaper if you were interested. An inverter/charger would work also, and you could use that with the Super Start battery you already have. They're basically UPSs, except they're BYOB (bring your own battery). They're a bit expensive though, and have some of the same drawbacks as UPS.

 

[kswan]

Thanks for all the information. This is how I envisioned the wiring. I was going down this path because I had the free UPS and 1 battery already for the pump backup so all I would need is an additional battery. It seems that I need to do some more research. Thank you again for all the information.

 

[chipmunkofdoom2]

@kswan I see what you mean now.

From purely an electrical perspective this should work. You would need to open the Ace in the Hole and wire the terminals directly to your second battery's terminals. You would also not want to use the battery maintainer, as there's no need for it: both the Ace in the Hole and the APC have chargers and both will attempt to charge the batteries. The Ace in the Hole would then have 12VDC as intended, and the UPS would have 24VDC.

One concern is that I'm not sure how these devices are going to behave when the sense the charge voltage from one another. I don't think anything dangerous would happen (although it could), but I would at the very least expect them to not charge properly. APCs in particular are very picky with regards to charging and battery voltages. If the APC senses the Ace in the Hole attempting to charge the battery while it attempts to charge both batteries, the APC may refuse to work at all.

The other concerns it that the battery on your sump pump will be drawn down more frequently than the other battery. This will cause uneven wearing and uneven charging. How much this affects the system depends upon how much you discharge the batteries and how often the sump pump runs.

This is a messy solution. I would personally not attempt to use the sump pump device (the Ace in the Hole). It might work if the APC charger and the Ace in the Hole play nice. I don't think that's a given though. You're adding a lot of variables to the system that neither the UPS or the sump pump designers ever expected. It could, in theory, work.

 

[kswan]

Thanks again @chipmunkofdoom2! I really appreciate the amount of feedback. The sump pump has yet to run under battery power since I put it in last spring. I bought it after the sump got really full during a prolonged outage with heavy spring rain. I'll do some thinking and decide if I want to give it a try. Either way I'll follow up with the result.

 

[RamsReef]

Very well done sir. Thank you for putting that all together. This may very well save someones tank in the near future. I have two APC computer battery back up boxes connected only to a critical intank flow device in each tank. I also have an old ECOTEC battery back up box I intent to cut apart and utilize a larger battery to back up my ecotec pumps because they can utilize 12 volt circuits very simply. You have resparked my interest in this project.
Folks should also consider the usable life of there battery is 4-6 years and should put it on a replacement schedule and possiblly keep it inside a small plastic vented tub. I have seen a hand full of old batteries on trickle chargers go a bit haywire when they fully melt down. Usually only over working the trickle charger. But its not to hard to set in into something to stop any splits if a battery case splits then the plates fail inside eventually.

Get wet cells.

 

[Johnny M]

Very good article and following discussion. Don't have the skill set to do that kind of electrical work. Right before Sandy ( live in Central NJ) bought a portable generator - 8500W - even had electrician put in an outside to inside dual outlet so I didn't have to keep a window slightly open to have the power cords come in. I can keep everything on my tanks going, plus refrigerator , basement sump, heater or small window a/c, depending on season & a few lights. Only downside is gas tank on gen. has to be filled every 6 hours.

 

[chipmunkofdoom2]

So the past few weeks I've been ordering parts to build a better transfer switch. The old one worked just fine, but it was not the prettiest thing. I made the new one for probably around $35. Here are the parts I used:

- generic project box: $10
- DPDT 120VAC relay: $10
- IEC320 C14 sockets: about $1 each
- Outlet with wall plate: < $5 at hardware store
- Various connectors (ring terminals, .187" quick disconnects): < $5 at hardware store
- Various machine screws and nuts: < $5 at hardware store
- Assorted wire

Construction was pretty straight forward. First, I cut holes in the project box for the IEC sockets:




Next, I cut a hole in the top for the socket:



Next, I secured the IEC sockets and the outlet with the machine screws. After that, I did a lot of soldering. I needed a quick disconnect on each side of each wire going from the socket to the relay. I also needed a quick disconnect on the two ground wires, although they would end up being tied together.



From the relay to the outlet, I used quick disconnects on the relay side and ring terminals on the outlet side. This allowed me to tighten down the wires to the outlet a bit more securely, since I could put the screws through the ring terminals:



Finally, put the box together and screw down the lid:



The utility power goes in the left socket, the inverter plugs into the right socket, and your outputs plug into the outlet. Pretty straight forward. I'm considering upgrading the unit to have a red LED to indicate power failure status, but that's a bit down the road. For now, it works, and it's a heck of a lot prettier than my previous attempt at a transfer switch.

 

[chipmunkofdoom2]

Very good article and following discussion. Don't have the skill set to do that kind of electrical work. Right before Sandy ( live in Central NJ) bought a portable generator - 8500W - even had electrician put in an outside to inside dual outlet so I didn't have to keep a window slightly open to have the power cords come in. I can keep everything on my tanks going, plus refrigerator , basement sump, heater or small window a/c, depending on season & a few lights. Only downside is gas tank on gen. has to be filled every 6 hours.

Thanks, I appreciate the feedback

Generators are a fine option. After a few days of being out of power, unless you seriously overbuild your battery bank, even with this project you'll likely need a generator. The only things to keep in mind with generators are the annual maintenance they require and that most don't turn on automatically. If you're not going to be around your tank to get to it quickly in the event of an outage, a battery backup might still be beneficial.

 

[Ranjib]

So the past few weeks I've been ordering parts to build a better transfer switch. The old one worked just fine, but it was not the prettiest thing. I made the new one for probably around $35. Here are the parts I used:

- generic project box: $10
- DPDT 120VAC relay: $10
- IEC320 C14 sockets: about $1 each
- Outlet with wall plate: < $5 at hardware store
- Various connectors (ring terminals, .187" quick disconnects): < $5 at hardware store
- Various machine screws and nuts: < $5 at hardware store
- Assorted wire

Construction was pretty straight forward. First, I cut holes in the project box for the IEC sockets:




Next, I cut a hole in the top for the socket:



Next, I secured the IEC sockets and the outlet with the machine screws. After that, I did a lot of soldering. I needed a quick disconnect on each side of each wire going from the socket to the relay. I also needed a quick disconnect on the two ground wires, although they would end up being tied together.



From the relay to the outlet, I used quick disconnects on the relay side and ring terminals on the outlet side. This allowed me to tighten down the wires to the outlet a bit more securely, since I could put the screws through the ring terminals:



Finally, put the box together and screw down the lid:



The utility power goes in the left socket, the inverter plugs into the right socket, and your outputs plug into the outlet. Pretty straight forward. I'm considering upgrading the unit to have a red LED to indicate power failure status, but that's a bit down the road. For now, it works, and it's a heck of a lot prettier than my previous attempt at a transfer switch.

Very nice

 

[Finatik]

Excellent post @chipmunkofdoom2 !!! I live in an apartment which means a generator is not an option for me, so for the past few months, I have been trying to gather all of the supplies and components I need to set-up an emergency back-up system such as you describe. I have two 120 gallon tanks, so I purchased two of these:

https://www.amazon.com/Renogy-2000Watt-Inverter-Charger-Outlets/dp/B01MUCN1WJ/ref=redir_mobile_desktop?_encoding=UTF8&keywords=2000w power inverter&pi=AC_SX236_SY340_QL65&qid=1490971311&ref_=mp_s_a_1_39&sr=8-39

I also have two VorTech battery backups to power two sets of MP40s so that I will be able to maintain aeration and circulation in both tanks in the event of a prolonged power outage.

Next, I am planning to purchase 4-6 Deep Cycle batteries to go along with the Inverters I purchased so that I will be able to power the four 250w heaters, and all of the pumps I have between my two 120 gallon tanks.

BUT... I am not an Electrician, so this stuff is kind of foreign to me and I don't really know if my understanding of all of this is correct. Nor do I know if I've thought of ALL of the different pieces and parts I'll need to make this actually work. This is just what I've come up with from doing my own research about how to set this all up. So what are your thoughts @chipmunkofdoom2 ? Do you think this set-up will be able to keep my two tanks running for 24-48 hours in the event of a prolonged emergency power outage ? With hurricane season fastly approaching, I want to get this all set-up so that I don't have to worry about it any more. *** Any advice or suggestions you can give me about this would be greatly appreciated ***

 

[mckinleyw]

Can i get input on a few parts Ive been looking at? I am looking at an overkill so it wont be a cheap list of parts. these 2 items seem to have all i need.

first battery... 200ah 12v $400
https://www.amazon.com/gp/product/B079J4QSC6/ref=ox_sc_act_title_1?smid=A1Y5M6NIMLYBIZ&psc=1

then i found this inverter... $500
https://www.amazon.com/gp/product/B076Y9J512/ref=ox_sc_act_title_2?smid=A2FRZPVYLUOBKW&psc=1

thoughts??

 

[siggy]

Nice work

 

[chipmunkofdoom2]

@Finatik Inverter/chargers are not my personal preference, but they will certainly work just fine for this application. It's really all you will need beside a battery. It incorporates all the parts you need to make this system work. I would only order 1 inverter/charger, unless you plan to have a separate battery/battery bank for each. Since each inverter/charger will try to charge the batteries when the power is on, you may get some unexpected results from two chargers trying to run at once. Unless they're dumb float chargers and all they do is float the batteries to a certain voltage. But unfortunately, there's no way to know what mode of charging these things will use.

The batteries will be the real deciding factor in how long your system runs. Ideally, you should get a Kill-A-Watt to measure how much power all the components you want to run will draw. Because you're talking about a rather expensive system, I personally would not try to plan capacity without knowing exactly how much power you need. If you use Apex controllers, you might be able to tell your power usage from the controller.

If I can talk you out of running the heaters on the battery backup, I would. Unless you live in a cold area, the tanks will likely not cool down very much. Additionally, heat is not usually an emergency unless we're talking about several days of power loss and very cold temperatures (in the 60*s). There's no reason you couldn't run the heaters on batteries, but it's going to eat a lot power.

I've created a calculator to make finding the power requirements of battery backup systems easier. It can be found here. Simply plug in the wattage of your equipment, how many hours you want it to run, and the calculator will give you the amount of amp hours you need. The inefficiency and discharge limits can be changed, but those two values are good estimates. If we use 100W as your load and a 24 hour run-time, the calculator says you need 275 Ah of battery power. This battery is 100Ah, so you would need three of them in total (100Ah * 3 = 300Ah) to run 100W for 24 hours. You'd need to double the batteries for double the run time. Conversely, you could double the run time by reducing the load in half (running only 50W of equipment instead of running 100W).

If we assume your 1000W heaters only run half the time, they would increase the amount of batteries you need by roughly six times. The costs really start to add up quickly if you want 24-48 hours of run time. In my case, I run a 7W pump in my sump and a 7W pump in my display (a Koralia 425 and Jebao PP-4), so a single 100Ah battery will give me 60+ hours of runtime. You can certainly buy as many batteries as you can afford and this system will work. But at a certain point, a solution like Tesla's Powerwall may actually be more economical.

 

[chipmunkofdoom2]

Can i get input on a few parts Ive been looking at? I am looking at an overkill so it wont be a cheap list of parts. these 2 items seem to have all i need.

first battery... 200ah 12v $400
https://www.amazon.com/gp/product/B079J4QSC6/ref=ox_sc_act_title_1?smid=A1Y5M6NIMLYBIZ&psc=1

then i found this inverter... $500
https://www.amazon.com/gp/product/B076Y9J512/ref=ox_sc_act_title_2?smid=A2FRZPVYLUOBKW&psc=1

thoughts??

Both these parts would work, but I would suggest a few modifications.

First, the inverter. While it will work, I've not heard of that brand (AIMS) before. Since you likely won't need a ton of wattage in an aquarium battery backup scenario, I would personally get a smaller wattage inverter from a known brand, like Tripp-Lite, Renogy or Xantrax. They're all about the same price as the AIMS, but these are all known entities in the electronics and inverter world.

Second, the battery. That battery will also work, but it's a gel cell. These batteries require very specific charge profiles to remain operational, and they likely will not last very long on an inverter/charger (unless it's a very high-quality inverter/charger). At the very least, the inverter would have to explicitly state that it supports gel batteries, and even then, I would try to call or email the manufacturer to make sure. I would personally stick with AGM batteries. This 100Ah model is about $170 at the time of this post.

 

[mckinleyw]

Both these parts would work, but I would suggest a few modifications.

First, the inverter. While it will work, I've not heard of that brand (AIMS) before. Since you likely won't need a ton of wattage in an aquarium battery backup scenario, I would personally get a smaller wattage inverter from a known brand, like Tripp-Lite, Renogy or Xantrax. They're all about the same price as the AIMS, but these are all known entities in the electronics and inverter world.

Second, the battery. That battery will also work, but it's a gel cell. These batteries require very specific charge profiles to remain operational, and they likely will not last very long on an inverter/charger (unless it's a very high-quality inverter/charger). At the very least, the inverter would have to explicitly state that it supports gel batteries, and even then, I would try to call or email the manufacturer to make sure. I would personally stick with AGM batteries. This 100Ah model is about $170 at the time of this post.

THANK YOU. I was looking at those as well. I wasnt sure about the gel or sealed lead acid. Im not sure on my actual usage. My main concern is getting too hot. I live in florida. Is there anything about running a chiller on a bat. backup? i know the surge on start up and stuff is hard to deal with. I will only need for when power outage before i can get genny out and running.

 

[chipmunkofdoom2]

THANK YOU. I was looking at those as well. I wasnt sure about the gel or sealed lead acid. Im not sure on my actual usage. My main concern is getting too hot. I live in florida. Is there anything about running a chiller on a bat. backup? i know the surge on start up and stuff is hard to deal with. I will only need for when power outage before i can get genny out and running.

You're welcome, glad to help

A chiller will be tough to run. To start with, since we're talking about a device that uses a compressor, I would definitely get a pure sine wave inverter. Modified sine wave might be fine for aquarium pumps that are submerged in liquid, but with something as expensive and complex as a chiller, I would want pretty clean power going into it.

There are two big issue here that will need addressing. The first is the start-up current. As you mentioned, it's probably going to be pretty high with a compressor. Some refrigerators draw up to 13A on startup, which is almost 1600W at 120VAC. Most modern refrigerators use about 100W - 200W when running, so the inrush current in this case is huge. There aren't many good ways to measure start up current without a high-quality multimeter, like a Fluke with an INRUSH mode. I don't think a Kill-A-Watt will tell you in-rush current. You could switch to the WATTS mode and hope you see the inrush on the meter, but this may not work if the start-up is fast. You might be able to call the chiller manufacturer and ask how much inrush current they expect with that chiller. You could also just skip all that and get a really high capacity (2,000W+) inverter. But, if you're going with a PSW inverter, this is going to get costly fast. Plus, if you undersize the inverter, it won't run your load at all. This 2,000W model from Renogy has a 4kW surge rating. If your chiller is 400W or smaller, this allows for a 10X current draw at startup, which I think is reasonable. Any larger than 400W though might be pushing it.

I'm honestly not sure how I would handle this one. Measuring inrush current with a multimeter is probably the most accurate way to do it, but it requires messing around with AC voltage and requires a good multimeter. @Brew12 has a lot more electrical experience than I do. @Brew12 do you have any thoughts on how best to measure the inrush current for a chiller?

The other big issue is battery power. So assuming we've dealt with the large inrush of current, we now have to power the thing. A lot of chillers use over 300W. Some of the 1/2HP models report using up to 7A (that's 840W at 120VAC). If we split the difference and assume the chiller uses around 500W and you have 50W in pumps you want to run, that's still a huge load. After accounting for inefficiency and DOD, that's 252 amps to run the load for 4 hours. You would need at least 3 of the battery I put in my previous post to run this system for just 4 hours. That does assume the chiller will be running 100% of the time, which might not be the case. But, if your AC is off and the house is heating up, the chiller might just be running continuously.

I think the best place to start is getting a Kill-A-Watt and measuring how much power your equipment uses. Try measuring how many watts the chiller draws while it's on so that you can estimate that it's running 100% of the time.

 

[mckinleyw]

You're welcome, glad to help

A chiller will be tough to run. To start with, since we're talking about a device that uses a compressor, I would definitely get a pure sine wave inverter. Modified sine wave might be fine for aquarium pumps that are submerged in liquid, but with something as expensive and complex as a chiller, I would want pretty clean power going into it.

There are two big issue here that will need addressing. The first is the start-up current. As you mentioned, it's probably going to be pretty high with a compressor. Some refrigerators draw up to 13A on startup, which is almost 1600W at 120VAC. Most modern refrigerators use about 100W - 200W when running, so the inrush current in this case is huge. There aren't many good ways to measure start up current without a high-quality multimeter, like a Fluke with an INRUSH mode. I don't think a Kill-A-Watt will tell you in-rush current. You could switch to the WATTS mode and hope you see the inrush on the meter, but this may not work if the start-up is fast. You might be able to call the chiller manufacturer and ask how much inrush current they expect with that chiller. You could also just skip all that and get a really high capacity (2,000W+) inverter. But, if you're going with a PSW inverter, this is going to get costly fast. Plus, if you undersize the inverter, it won't run your load at all. This 2,000W model from Renogy has a 4kW surge rating. If your chiller is 400W or smaller, this allows for a 10X current draw at startup, which I think is reasonable. Any larger than 400W though might be pushing it.

I'm honestly not sure how I would handle this one. Measuring inrush current with a multimeter is probably the most accurate way to do it, but it requires messing around with AC voltage and requires a good multimeter. @Brew12 has a lot more electrical experience than I do. @Brew12 do you have any thoughts on how best to measure the inrush current for a chiller?

The other big issue is battery power. So assuming we've dealt with the large inrush of current, we now have to power the thing. A lot of chillers use over 300W. Some of the 1/2HP models report using up to 7A (that's 840W at 120VAC). If we split the difference and assume the chiller uses around 500W and you have 50W in pumps you want to run, that's still a huge load. After accounting for inefficiency and DOD, that's 252 amps to run the load for 4 hours. You would need at least 3 of the battery I put in my previous post to run this system for just 4 hours. That does assume the chiller will be running 100% of the time, which might not be the case. But, if your AC is off and the house is heating up, the chiller might just be running continuously.

I think the best place to start is getting a Kill-A-Watt and measuring how much power your equipment uses. Try measuring how many watts the chiller draws while it's on so that you can estimate that it's running 100% of the time.

thanks again very good info here just got a killawatt last week, will be checking a few of these wattages. I may just end up using frozen waterbottles again when needed. Also is there any online data on some of the more popular equipment electricity usages?

 

[Brew12]

You're welcome, glad to help

A chiller will be tough to run. To start with, since we're talking about a device that uses a compressor, I would definitely get a pure sine wave inverter. Modified sine wave might be fine for aquarium pumps that are submerged in liquid, but with something as expensive and complex as a chiller, I would want pretty clean power going into it.

There are two big issue here that will need addressing. The first is the start-up current. As you mentioned, it's probably going to be pretty high with a compressor. Some refrigerators draw up to 13A on startup, which is almost 1600W at 120VAC. Most modern refrigerators use about 100W - 200W when running, so the inrush current in this case is huge. There aren't many good ways to measure start up current without a high-quality multimeter, like a Fluke with an INRUSH mode. I don't think a Kill-A-Watt will tell you in-rush current. You could switch to the WATTS mode and hope you see the inrush on the meter, but this may not work if the start-up is fast. You might be able to call the chiller manufacturer and ask how much inrush current they expect with that chiller. You could also just skip all that and get a really high capacity (2,000W+) inverter. But, if you're going with a PSW inverter, this is going to get costly fast. Plus, if you undersize the inverter, it won't run your load at all. This 2,000W model from Renogy has a 4kW surge rating. If your chiller is 400W or smaller, this allows for a 10X current draw at startup, which I think is reasonable. Any larger than 400W though might be pushing it.

I'm honestly not sure how I would handle this one. Measuring inrush current with a multimeter is probably the most accurate way to do it, but it requires messing around with AC voltage and requires a good multimeter. @Brew12 has a lot more electrical experience than I do. @Brew12 do you have any thoughts on how best to measure the inrush current for a chiller?

The other big issue is battery power. So assuming we've dealt with the large inrush of current, we now have to power the thing. A lot of chillers use over 300W. Some of the 1/2HP models report using up to 7A (that's 840W at 120VAC). If we split the difference and assume the chiller uses around 500W and you have 50W in pumps you want to run, that's still a huge load. After accounting for inefficiency and DOD, that's 252 amps to run the load for 4 hours. You would need at least 3 of the battery I put in my previous post to run this system for just 4 hours. That does assume the chiller will be running 100% of the time, which might not be the case. But, if your AC is off and the house is heating up, the chiller might just be running continuously.

I think the best place to start is getting a Kill-A-Watt and measuring how much power your equipment uses. Try measuring how many watts the chiller draws while it's on so that you can estimate that it's running 100% of the time.

Thanks for the invite! And this is a nice little system you have come up with, I love how it is both safe and reliable. Great job! I know you said you test in monthly, any thoughts on adding a push button, momentary, normally closed, in line with the relay coil to make testing easier?

Chillers and batteries aren't going to work well together and you have identified both reasons why. Chiller inrush can be 7x to 10x the running current but that is when powered from a normal power supply. When powered from a batter, the starting surge can cause a voltage drop on the DC side which causes even more current to flow. I wouldn't try this unless I had the space to install something like 4 to 6 battery bank of boat trolling motors. Not to mention the high capacity inverters.

I would consider a cooling fan blowing across your sump as a better option until you can get your chiller up on the generator. You could even use something as elegant as what @chipmunkofdoom2 put together with a little modification. When normal power is available, the relay provides normal power to the chiller. If normal power is lost and the system switches to battery, the chiller is left off and the cooling fan turns on.

 

[mckinleyw]

Thanks for the invite! And this is a nice little system you have come up with, I love how it is both safe and reliable. Great job! I know you said you test in monthly, any thoughts on adding a push button, momentary, normally closed, in line with the relay coil to make testing easier?

Chillers and batteries aren't going to work well together and you have identified both reasons why. Chiller inrush can be 7x to 10x the running current but that is when powered from a normal power supply. When powered from a batter, the starting surge can cause a voltage drop on the DC side which causes even more current to flow. I wouldn't try this unless I had the space to install something like 4 to 6 battery bank of boat trolling motors. Not to mention the high capacity inverters.

I would consider a cooling fan blowing across your sump as a better option until you can get your chiller up on the generator. You could even use something as elegant as what @chipmunkofdoom2 put together with a little modification. When normal power is available, the relay provides normal power to the chiller. If normal power is lost and the system switches to battery, the chiller is left off and the cooling fan turns on.

THANK YOU @Brew12 for the suggestions.