Hey guys. Ultrasonic on the main wet well at the pump station, 4-20mA back to a 1769-IF8, and for three weeks the reading has been jumping. Sits at 1.4m, next scan it's 3.6m, then back down. Lead pump starts and stops on those spikes and it did eleven starts in an hour on Tuesday, which is going to cook a motor sooner or later.
Meter on the loop at the card says the milliamps jump the same way the tag does, so it isn't the PLC end. Cleaned the face of the transducer with a rag and it was fine for a day, then back to the same thing. Read a couple of threads where people said bad screen on the cable, so I swapped the run back to the panel for a spare screened pair, no difference there either.
Level was rock solid for four years. Is it just a worn out sensor, or is there something about a wet well that eats these?
Four good years then three bad weeks means something changed, not something wore out.
Where exactly is it mounted, and how far is the face off the high level? One photo looking straight down from the transducer would tell us more than the numbers will. And has anything gone in the well lately, guide rail, new riser, a float bracket?
It's on a bracket off the access cover, maybe 300mm in from the wall, face about 4m above the slab so roughly 2.6m of clear air at normal level. Photo of the well attached, it's not pretty down there.
Nothing new in the well that I know of. What I can tell you is the spikes are worst after the works has been washing down, and there's a lot of foam sitting on the surface when it's bad. Never joined those two up until you asked.
Meter the loop again with the well quiet and then again when it's spiking, but do it at the sensor head, not at the card. If it swings at the head too then the transducer is making the bad number and you're done arguing.
While you're down there check the earth. I reckon a 4-20 sensor on a bonded lid next to a VFD panel picks up all sorts, and it shows up as exactly this kind of jump. I've had two of those.
Metered it at the head while it was spiking. Same swing, 6mA up to about 13mA and back, so the sensor is genuinely sending that. Lid bond is good, 0.3 ohms back to the panel bar, and the drive panel is 12m away on its own conduit.
So the sensor is telling the truth about what it can see. Stuck on what it thinks it's seeing.
Foam. That's your answer, and the wash down is why it started.
An ultrasonic works on a returned echo. Foam soaks the pulse up so the head hears nothing off the surface, and then it grabs the next loudest thing in the cone, which at 300mm off the wall is the wall, a bracket, or condensation on its own face. That's your 3.6m. Wiping the face bought a day because you took the film off.
Get it on a clear shot over open water, and fit a stilling tube so the cone can only ever see one thing. If the foam is heavy every wash down, a submersible transducer is the honest answer. Filter in the PLC after that, not instead of it. Wet well level, there's a writeup here: https://plctr.com/utilizing-plc-in-water-and-wastewater-treatment-systems/
Right, it was the foam, and the mounting was making it worse. We moved the bracket out to 900mm off the wall over open water and dropped a 100mm PVC stilling tube down past low level with a vent hole near the top. Reading has been flat since we finished, one wash down included, and the lead pump is back to three starts an hour.
Put a 5 second average on the analog in the program as well, which I'd have gone straight to if nobody had explained the echo side of it. What I still don't get is why four years of wash downs never did this, unless somebody changed the detergent. Cheers pat_k, edj and carlh, saved me buying a sensor that was working fine.
Stilling tube is worth fitting on every well. Wish more people did.