A whole lot of good ideas and questions that need to be answered:
1) The easiest way to wire this is two sump pumps... one driven off AC power, the second (backup) driven from the battery. With power on, the system is redundant. With power off, there's a single point of failure (backup pump that is seldom... if ever... used, unless tested)
2) Note that I plan on using sump pumps that have floats... so even when power is being "routed" to one sump pump (we'll call it the 'hot' pump), it doesn't actually run unless the float completes the circuit
3) The inverter and relays are essentially a "poor-mans" automatic transfer switch. When power fails, stop powering from the GRID and power from the inverter instead
One is if the pumps are normal 120 VAC you should only "switch" the hot leg and not the neutral.
Easily done, but then I wouldn't be able to switch between inverter and house_AC, right? I don't think I can have the inverter hot be serviced with the house neutral or vice versa.
EDIT: Sorry, I think I get what you mean. In the pump diagram, connect all the neutrals together. These neutrals would be contacted by the other relay to either the inverter_AC or the house_AC. I guess I could use either a SPDT switch, or a DPDT switch while only connected the hot to it, right?
Also, I'm not sure what logic you would use to switch the relays because one pump would always be on.
One pump would always be "powered" but the float would control whether the pump was on. I'm sorry for omitting this detail.
Can you describe a little more how you are planning on switching the relays (and what coil voltages you will use for this)?
It depends on what relays I acquire. Alas, I cannot use the ELK-924s. Anyway, as I understand it, DPDT relays are such that one pair of poles is contacted if there's no power across the relay, and the other pole is contacted if power is applied to the relay. I was planning on using a wall-wart matched to the voltage and AC/DC type. For the house_AC and inverter_AC lead, I'd wire the inverter_AC to the contact that's connected when there's no voltage, and the house_AC to the other contact. In this manner, as long as there's power to the wall-wart, the relay would be connected to the house_AC side. If power is lost, then the relay would de-energize and contact would switch to the inverter_AC source.
For the DPDT that controls the pumps, I was leaning on using the ELK to round-robin the relay from left to right with every operation. I was planning to wire in an additional float that located higher in the sump pit than the two sump pumps. If this float contacts, than there's a pump failure, and the ELK should immediately change to a different pump. Alternatively, I guess I could build a simple circuit board that does the same.
For your DC question, I would not tie the negative of the 12 volt system to earth ground as this is a signal patch (i.e. not like an AC ground). Also, I would only switch the hot leg of the input to the inverter as well.
The inverter has a separate grounding screw to connect to the chassis, I guess I never thought that this was the negative.
I can switch only the hot leg of the inverter input. I assume thats the "positive" cable from the battery?
You might be overthinking this a bit. If you want to have two pumps, I would give them each a float switch (to cover the possibility of a float switch failure) and wire the two of them so that they both have power all the time. Then I'd simply install one float switch a bit lower than the other. That way the pump with the lower float switch would be the primary pump and run most of the time, but if that pump (or it's float switch failed) then the water would rise an inch or two and trigger the float switch on the backup pump.
It might be a bit over-engineered
Your design has merit, too.. but I think there are two disadvantages:
1) You only know that the backup pump works if you test it. This is a bit of pain, since the only way to do that is manually manipulate the float. I'd prefer to provide an airtight seal around the sump pit to further reduce my radon levels... which would make regular tests annoying... The above design means that every other time your sump activates, you'd be testing the backup device.... automatically.
2) I'm using a good, inexpensive, but relatively low powered inverter. In battery-operated mode, its conceivable that the "lower" pump fails in an on-state -- eg: air lock, broken impeller, leak in the outflow pipe, etc -- and when the backup turns on, I'd double the load on the inverter, and in fact, likely exceed its rated ability.
It would be a good idea to test the backup pump and float switch periodically as they should never trigger if things are working well, but I think that this would eliminate a lot of complication and also eliminate your relay system as another single point of failure (what happens if one of the relays fails?)
agreed!