1

I'm trying to replicate the battery backup setup described in this video and in text below. I'll state what the video demonstrates, then share the parts and plan I have. I'm looking for input on how correct and safe my plan is.

In the video I based this on, someone DIY's an inverter-charger UPS type of power backup system, which has benefits of lower total cost and easier to maintain (you can replace components rather than entire system). Since I have some components already, this is preferable.
In the video they connect a PROwatt SW inline transfer relay to 120VAC household circuit sufficient for the sump pump, and to a 12VDC-120VAC inverter, then on the other end of the transfer relay they have the sump pump. The inverter is connected to a 12VDC battery, which is connected to a charger plugged into 120VAC.

Here's my plan:

  • There's a 20A 120VAC circuit for a sump pump, with one duplex receptacle on it.

  • I have a 12VDC LiFePO4 battery and plan to expand capacity with more of them in parallel, and I have a compatible Victron BlueSmart IP65 Charger (12 VDC, 15A). The battery would connect to the charger, charger plugged into the 20A 120VAC household circuit. That's my DC power supply, which I figure the Smart Charger can always keep topped off and not over-charged so long as mains power permits, while also safely isolating the DC power from the mains power supply.

  • Then I'll get a 2000W pure sinewave inverter with GFCI receptacles to supply the 900W sump pump (rated at n more than 7.6A continuous draw at 120VAC). I expect the inverter, battery, and charger would always be 'on' and simply not drawing much from the battery while keeping it topped off, so long as AC power is on. My understanding is there should also be a 15A fuse between battery and inverter, and there should be a ground from the inverter to the house's ground copper.

  • To put the sump's load on the inverter+battery only when AC power fails, I'd use a transfer relay. Sump plugs into transfer relay, transfer relay plugs into house's AC and into inverter's DC power. This would act as a switch so that no load is on the inverter unless a load is generated that AC power cannot supply. The PROwatt SW model linked to above seems fine. That said, I'm not experienced in DC to AC connections (I have decent familiarity with AC and amateur familiarity with DC systems).

I understand the DC leg of this system is rated to 15A because the charger, inverter, and transfer relay are all rated for 15A. For that reason, I figure I need a 15A external fuse between the battery and the inverter on the hot wire to make sure the load on the battery does not exceed that amount. (The inverter I'm looking at has 15A GFCI receptacles and an internal fuse, as well.)

For the AC leg of the system, I don't see a reason why higher amps rating would be a problem. The 20A duplex receptacle and circuit (wired with 12/2 cable and 20A breaker) should be able to handle any 15A loads coming from the DC system (via the charger's load), and the 20A circuit was designed to be more than enough for the sump pump load directly.

Are the parts and ratings I described all compatible? Is there any parts missing or areas of concern I didn't address?

isherwood
  • 158,133
  • 9
  • 190
  • 463
cr0
  • 1,716
  • 4
  • 34
  • 59

2 Answers2

3

This is not a good plan, for many reasons:

  1. This is going to be at least as expensive (maybe significantly more expansive) and a lot more effort than simply buying an off-the-shelf battery-backed sump pump (which will also have a warranty as a system).

  2. You'll need to ensure that the charger won't drain the battery if it is left connected when the AC power goes out. It'll probably be fine, but it's not guaranteed.

  3. Inverters do have significant idle draw. Probably not enough to worry about the cost when the battery is being charged by the grid, but definitely enough to waste a significant chunk of your battery capacity if the grid power is out for a day or more.

  4. A 2000W inverter may or may not be enough to start a 900W motor. Motor can draw many times their rated power for a short time when starting. It's difficult to predict whether any given motor/inverter pair will work, without trying.

  5. Your conceptions about the DC side current are incorrect. The DC side (between battery and inverter) will not be operating at 15A with a 900W load. 900W from 12V requires 75A, plus some more for inverter inefficiency. If the motor requires the full 2000W of the inverter to start, that will be a peak of 167A.

  6. Paralleling batteries for extra capacity is not a good idea. Batteries will age and discharge differently. They need to be charged independently. They should be isolated from each other (e.g. with diodes) so that if one goes bad, the good one doesn't simply discharge into the bad one when the power goes out. Wiring cells or batteries in series (either 6 2V cells for a 12V system, or two 12V batteries for 24V) can give more capacity but doesn't do much for redundancy - if one dies and you don't notice, the whole thing may still be unusable.

nobody
  • 7,701
  • 3
  • 36
  • 47
3

Native 24VDC & 12VDC bilge pumps are available (at quite reasonable cost) if you want a battery pump system you can DIY easily.

You get to skip the inefficiency of running an inverter, and the cost of the inverter. The battery charger keeps things charged when the power is, on, the battery keeps pumping until it gives out (or better, your low voltage cutout shuts it off, but some folks want to kill the battery for that last bit of pumping - either way the sump will overflow if power is out too long.) The pump does not know nor care if the power is on to the charger.

Ecnerwal
  • 235,314
  • 11
  • 293
  • 637