Anyone have experience with power sensing plugs? I'd like to have a way to monitor if I've got an AC sump pump cycling more often than expected. I took a stab at this a few years ago with a z-wave device and was less than happy with it. I'd be open to trying a zigbee solution.
I just want to know when the pump comes on and goes off. An added bonus would be being able to monitor if this device disappeared, as I've had one pump trip it's GFCI and I had no idea it was offline.
At some point, sure, I'd probably be better off with something that actually monitors the water level. Is there a plug-and-play option out there already? I do not want to get into lashing together something. My technical to-do list is already backlogged enough.
I don't need to know power consumption, but wouldn't reject being able to collect it. Nor do I need on/off control of the power, I'd just be using this purely to monitor whether the pump powered itself on.
I know this is only one aspect of your question, but it is the one I can help on, so I will answer it. I plugged an energy monitoring plug into the same gfci outlet that the sump pump is plugged into. Then, I have the following rule:
Basically, this rule runs once per hour and checks to see whether the energy duration has changes since last time the rule ran. If the device is not getting power because the gfci tripped, then it keeps returning the same old values. The rule doesn't detect a problem right away, but it detects it within an hour which is good enough.
I have the sump plug plugged into the opposite outlet as the ZWave plug. Every so often, the ZWave plug seems to turn off even though I didn't command it to. Even if it is turned off, it still continues to report as long as it is plugged into a working outlet.
You can get flood sensors that work with Hubitat. For sump pump monitoring, you put the flood sensor high enough that it wouldn't get wet unless the sump pump isn't working.
For sensing level (crudely) I have an Everspring flood detector that has a remote sensor on a fairly long cord. I tie-wrapped the sensor inside the sump at a level above where the pump normally drains it - so the sensor will only get wet if the pump fails resulting in the water getting too high. I like this model because it can be wall mounted and it uses "regular" batteries (3 x AA).
If your sump pump is controlled by a float of some sort. Can you use a contact sensor? Magnet on float mechanism and battery operated device connected to something appropriate.
Not a direct answer to OP but here is what I do for my sump pump monitoring:
Standard leak sensor placed next to the sump pit. I'd put it into the pit, but it's sealed for radon mitigation, though obviously if the sump pump failed the water would get past the sealing.
My backup sump pump is something like this. The control module has hookups for a basic two wire alarm. So I ran those into a standard zwave contact sensor. Now if/when the backup system turns on, I get an immediate notification via Hubitat.
What's with the 113.25 days variable? Just a random range chosen?
The outlet turning off is what I specifically want to avoid. I ran into some similar voodoo when I first poked at this idea with an Aeon z-wave gizmo a few years ago. I was likewise interested in monitoring my basement freezer, but the damned thing randomly turning itself off was a deal breaker.
Honestly, I'd really prefer something that was completely passive at monitoring the power consumption, but I've yet to see anything that does this. Given the potential for water damage from an inoperative pump and likely insurance investigation I want as little potential for screwups as possible. I don't want anything lurking around the sump pump that'd give anyone pause to deny coverage, if you catch my meaning.
it's in an sealed enclosed well (double-duty for radon venting). I open it yearly to check that it hasn't gotten clogged or otherwise inoperable. It's an integrated unit, no separate float or point where I could tap for a sensor.
I have two sumps, both fed from a french drain around the outside of the foundation.
The one on the uphill side is guaranteed to run after every rain. I can usually hear the water from that one. It wasn't until I heard the pump short cycling that I caught that the backflow valve had failed. It is plumbed such that it goes up, over and out to the downhill wall. Not a great design, but it is what it is and this being a finished basement level... isn't going to change significantly.
The good part is the french drain runs around the whole house, so if the high-side sump isn't able to handle it, the low side sump will get the overflow and eject it from there. It's the low side that randomly decided to trip the GFCI. I'd be in bad shape if the high side backflow failed and the low side GFCI had tripped. Thus my interest in monitoring.
I'd very much like to have a better way to monitor for that sort of cycling or excessively long run times that would indicate a potential failure. But I'd like to do it in such a way that it doesn't add on MORE layers of technical debt.
Kind of far out but might a sensor with a vibration capability work? I have a monoprice mct340 motion sensor. The motion capability seems pretty sensitive. Its a MCT340. I'm sure there are others.
The key is that the energy duration changes regularly as long as the device has power (even if it is switched off). My test in the rule looks for changes. It doesn't care about what the value is. That is why it needs to store the previous value in a local variable. This rule checks whether the gfci has tripped. It provides no direct information about whether or not the sump pump has run. (I can infer that the sump pump hasn't run if the gfci is tripped). It addresses that failure mode, but not other reasons that a sump pump might fail. Since I am actually monitoring the gfci, the sump pump doesn't need to be plugged into the ZWave outlet.
Installing a battery powered back-up sump pump is your best solution. Instead of telling you when there is a problem, it does something about it. It is still good to have something that tells you so you know to get the main pump fixed.
I have a Sense energy monitor at my house. It detected the sump pump pretty easily and tells me how often it runs. I put a rule in the Sense app to let me know if my sump pump hasn't run for 6 hours. At my house, that doesn't necessarily mean a problem if there hasn't been recent rain. However, that is a $300 solution.
Yeah, I've been eyeing the Sense service for ages, and detecting sump pump and refrigerator activity has always been at the top of my list for monitoring.
Does the sense setup let you see more active use than expected? That's my main concern, that it gets blocked again and keeps cycling itself over and over.
$300 is trivial compared to even the most basic of water damage remediation costs.
Oh, wait, maybe I'm missing something, how is it you have your devices connected?
Right now one of sumps is on a GFCI that has randomly tripped twice and has no other outlets on it's circuit. I suppose I could use the 2nd socket on this outlet and connect a plug device into it, and just monitor if that device disappears (indicating that it has lost power, presumably because the GFCI tripped and not a larger whole house outage). This wouldn't tell me anything about the pump's activity though.
I've tried some motion sensing and the false positives are almost as bad as the missed ones. There's too many other things in our house that could likely cause enough vibration to cause them to trip. I went down this road for some exhaust fan monitoring and discovered the fun of truck traffic, slammed doors and other outside factors causing false triggering.
Pressure switches, though, I have not looked into those. Not sure how effectively I could monitor the ejection pipe for pressure changes without introducing yet another maintenance item. It'd be less trouble to move to a different sump pump setup that had other contact closure options.
GFCI outlet has two sockets. The sump pump is plugged into one of them. The ZWave device is plugged into the other. I found that when the gfci trips, the ZWave device doesn't report a zero voltage. Instead, it continues to report whatever the last reading before the power failure was. Looking at energy duration was a way for me to detect that it kept reporting the same thing and therefore was offline.
As far as I know, the only notifications available in Sense are either a device being off for more than a specified time or being continuously on for more than a specified time.
Sump Pump failures are pretty serious. I've dealt with this issue too.
I use a ring range extender on a different outlet to report if my power goes out. Works very dependably.
I have my Sump Pump plugged in to a large UPS. It will run it for 30min-1 hr. That will buy me a little time. I have a 2nd Ring Range Extender plugged in to my UPS also. This tells me when my UPS fails also.
I use this Water leak sensor to tell me if my sump pump has failed. I put the sensor end just above where the water shouldn't reach in my sump pit. If it triggers, I know there''s a problem.
I have used a first gen Iris smart plug to tell me when the sump pump is cycling. It does work, and will report how much power the pump is pulling. To me, that piece is optional and I don't think it's currently in use. If something would switch it off- nightmare.
I recommend having a rule that runs on a regular basis, like every 30 minutes, that turns on the plug. If something unexpectedly turns it off, which sometimes happens, at least it won't stay off for very long.
I watched them for awhile after installing, and then determined peak wattage when the pump was operating, then watched for it to drop. Once it drops, thats counted as a pump event, and I add the event to a counter for daily cycles in a GV. Then each day, I sum the events.
I had intended to track this info in a databse with Node-Red/Grafana but...the daily counters is all I ever really got done...lol!
The Sump at the back of the house has operated 179 times today, and the front 64.
In the last 117 days, my pumps have run 20756 and 5353 time respectively.
I'm not measuring mean run time which might be useful....
