Update your browser to view this website correctly. Update my browser now
A typical pump station has a control panel above ground and below ground (called the wet well) there are two pumps and four float switches. The float switches are just tilt switches inside a plastic float. They start floating when the water level comes up to them and they turn themselves upside down. When the water level gets up to the second float from the bottom (the lead float), the control panel turns one pump on (which alternates between the two pumps each time). Then the pump runs until the water gets down to the bottom float (the stop float) and it turns off. If there is a problem with the pump or if the water is coming in faster than one pump can handle, the water will come up to the third float (the lag float) which will cause both pumps to come on. If the water still continues to come up higher, it will get to the high water alarm float which will usually turn on a red light on the panel. So the way it is now, if there is a problem and neither pump is running, someone has to see the red light (and hopefully the bulb is not blown), know what it means, and call me to come fix it, hopefully before the wet well overflows.
I'm using optocouplers to check the 120VAC power going through the control panel. The "L1" optocoupler will check the power coming out of the control power breaker and let me know if there is a power outage or a tripped control breaker. Some panels have a smaller fuse between the breaker and the floats, so I will use the "Ctrl" optocoupler to check for a blown fuse. Then the stop, lead, lag, and high optocouplers will check the power coming up from the four float switches to see their status. For example if the lead float is making contact but the stop float is not, I'll immediately know there is a problem with the floats. If the lead float and stop float are both making contact but a pump is not running, I'll know there is a problem with a pump immediately.
To see whether the pumps are actually running when they are supposed to be, I'm measuring current draw inductively from one of the three phases on each pump. I used the circuit design from the OpenEnergyMonitor project. An increase in the amperage going to the motor usually means there is trash wrapped around the impeller in the pump, so having constant amperage monitoring can detect other problems also.
I mainly added the Q1 and Q2 relays because a couple of the pump stations I check have variable-frequency drives to run the motors (mainly to convert single-phase into three-phase power). The VFDs will sometimes go into a fault mode when there is a surge or a power outage and they have to be reset. There are terminals on the VFDs to wire up external resets, which I plan on hooking up to the relays so I can reset them remotely. On "normal" stations I hope to wire the relays up to the motor starters to be able to have an override mode where I can control the pumps with the Core in case one of the floats quits working and the pumps won't come on with the control panel wiring.
For Internet access to send alerts when there is a problem, I found a company called Ting that will let you activate used Sprint devices on a prepaid plan. I bought a $20 Mi-Fi off eBay and I can use 100 MB for $9/month. I got the Mi-Fi on a very cold day last winter and I found out that it has a temperature sensor in it that will shut it off if it gets too cold (probably because of the lithium-ion battery). The only way to reset it after it shuts itself down is to disconnect power from it. So I added the Q3 relay and USB socket to my design so if the Core loses connection to the internet it can try to reset the Mi-Fi in case it has shut down due to temperature.