Last week’s discussions in our water treatment community focused on a range of technical challenges and educational needs. Members were particularly engaged with exploring operational issues like the unexpected behavior of VFDs during backwash. There was also a strong emphasis on the intricacies of PFAS sampling protocols, as well as a lively exchange on data quality assurance courses for PFAS. The rise in HPC despite improved turbidity stirred a lot of insights, prompting professionals to share their experiences and solutions.
This Week’s Hot Topics
VFD with stage fright at backwash
This thread dives into the peculiarities of variable frequency drives (VFDs) acting up during backwash cycles. It’s a common issue that can lead to inefficiencies if not addressed properly. Read more here
PFAS sampling SOP that passes scrutiny
Members are dissecting the steps to ensure sampling procedures meet regulatory standards, especially given the heightened scrutiny on PFAS. It’s all about getting it right the first time. Read more here
When turbidity drops but HPC rises
This discussion explores the puzzling scenario of dropping turbidity levels accompanied by an increase in heterotrophic plate count (HPC), a topic that has piqued interest for its potential implications on water quality. Read more here
Best CEU courses for PFAS data QA
Looking for the best continuing education units (CEUs) to enhance your PFAS data quality assurance skills? This topic compiles recommendations and reviews from fellow professionals. Read more here
Inline UV254 probes for SUVA control
Inline UV254 probes can be a game-changer for SUVA control. This thread discusses their application and effectiveness in real-world scenarios. Read more here
Need CEUs for PFAS and LCRI
Professionals are sharing leads on where to find CEUs that cover PFAS and Lead and Copper Rule Improvements (LCRI), a must-read for those looking to stay ahead in their certifications. Read more here
Thank you for staying engaged with our community. Your contributions and inquiries drive the conversations that help us all improve and innovate. Until next week, keep up the great work.
But we’ve had VFDs act up during backwash and solved it by disabling the PID and forcing a fixed 55 Hz for that step, with short accel/decel and a higher min speed — telling the drive “don’t overthink this.” If you still need control, use a setpoint ramp or feed‑forward with anti‑windup so the loop doesn’t chase the surge and trip like it’s trying to brake and accelerate at the same time. If anyone’s seeing nuisance overcurrent, also check the check valve and backspin before restart.
We cured jumpy backwash by coordinating the valve motion with the drive: when the backwash valve starts to open we assert a DI to the VFD to “freeze PID” for about 2 s, then release, and we slowed the valve stroke to about 4 s to soften the headloss step. If you can’t spare a DI, use the drive’s setpoint ramp or bumpless transfer feature. @Jordan it felt like trying to steer while someone kept yanking the wheel.
In our plant, the “unexpected behavior of VFDs during backwash” turned out to be the mag meter dropping to zero on empty‑pipe just as the valve cracked, which made the drive jump to max. We fixed it by setting the meter to hold last value for 3–5 s at backwash start — took five minutes in the menu and killed the spikes; small caveat: if you’re pressure‑based, this won’t help.
One thing that smoothed our backwash was enabling the drive’s S‑curve and switching to the ‘ramp set 2’ with about 3 s accel, plus a 25% min speed clamp until the mag shows stable flow; if the filter’s pressure‑sensitive, back off the clamp to about 15% or use a 2 s manual speed before handing PID back. @olivia2599 do you have S‑curve or a second ramp set on that VFD?
Quick example from a retrofit last summer — disabling the VFD ‘flying start’ only during backwash stopped the jump when it tried to re‑capture speed as the valve moved… If you rely on it for power blips, tie it to a backwash DI so it’s off for 5–10 s and re‑enables once the mag shows steady flow.
Building on @vtaylor, our fix was to freeze the PID during the valve stroke: a limit switch drives the VFD’s “integral hold” for about 4 s, then we release once the flow is steady. If you run backwash open-loop, bump the minimum frequency just above static head or you’ll still get a thump like a dropped toolbox.
Quick tip that saved us on a sand filter backwash: enable the drive’s “pipe fill” for about 2–3 s and block PID until the header hits about 8 psi, then hand off to flow control — , the hunting vanished. If your VFD doesn’t have that, add a 1 s input filter to the flow PV only during backwash, @danny_mills01; it tamed our nuisance HH alarms without touching the valve timing.
We killed a similar backwash lurch by dropping the VFD torque limit to about 60% while the valve is in motion and using the valve’s aux contact to ignore the flow signal until the stroke finishes; once the header settles, we restore normal limits. If you’re on a mag meter, add 1–2 s of transmitter damping or you’ll chase noise as the bed lifts. @vtaylor your “integral hold” note tracks with this — torque limiting kept the pump from trying to be a drag racer.
Had the same backwash lurch that was driving me nuts; the fix was disabling the VFD’s “flying start/speed search” only during backwash and switching to a 3–4 s S‑curve accel — zero hardware, 10‑minute tweak. If your pump can coast between cycles, keep speed search for normal runs but use a DI to select a “backwash” macro that kills it, @vtaylor. Anyone else run into this on ABB/Yaskawa drives?
On our dual-media filters, the jerk vanished once we enabled a “setpoint ramp” — we roll the pressure/flow setpoint up over about 6 s when the loop is enabled instead of stepping it, then let the drive take over, like easing onto the gas instead of stomping it… Small caution: if your mag meter is sleepy, add a 0.3–0.5 s PV filter or you’ll chase noise right at start.
Quick tip: we stopped the backwash kick by holding the PID integral (“integral hold”) and running the pump open‑loop at a fixed Hz for the first 2–3 s, then doing a bumpless transfer to auto; that kept the loop from unloading and slamming. Building on @ihill47, a tiny pre‑open of the wash valve (10–15%) before the drive starts also helps, like easing onto the gas instead of dropping the clutch. Caveat: if your transmitter has heavy damping, shorten it or add a brief alarm inhibit, or the handoff lags and the bump creeps back.