Wet Well Level Control

The band

Everything about wet well control follows from the vertical distance between where pumps start and where they stop. That band determines the volume moved per cycle, which determines run length, which determines both starts per hour and detention time.

Run time is roughly that volume divided by the net pumping rate, which is pump capacity minus inflow. The worst case for short cycling is low inflow, because the well fills slowly but the pump empties the band just as fast as ever. The worst case for detention is also low inflow, because the well sits full between cycles. This is why setpoints should be checked against the lowest expected flow, typically the early morning minimum, rather than against design flow.

Signal validation

The dangerous level failure is not one that reads zero. It is one that reads a believable number that happens to be wrong. Three checks in the controller catch most of these.

Range check

Compare the raw analog value against the NAMUR fault thresholds. Below 3.6 mA or above 21 mA is a device fault, not a level. Below the physical bottom or above the physical top of the well is also impossible and should be flagged.

Frozen value check

If the level has not changed by more than a small tolerance for longer than any real operating condition would allow, the reading is stale. A wet well level that has not moved in thirty minutes while a pump is running is not a level, it is a stuck reading.

Rate of change check

A wet well cannot change several feet in one scan. A jump beyond a physically possible rate is a signal fault or interference, and should be rejected rather than acted on.

Raw_Fault    = Raw < Count_3p6mA  OR  Raw > Count_21mA
Range_Fault  = Level < Well_Bottom  OR  Level > Well_Top
Frozen_Fault = |Level - Level_30min_ago| < 0.05 ft  AND  Any_Pump_Running
Rate_Fault   = |Level - Level_last_scan| > Max_Physical_Rate

Level_Bad = Raw_Fault OR Range_Fault OR Frozen_Fault OR Rate_Fault

IF Level_Bad THEN
    Level_Used = Last_Good_Level      // held, and marked as held
    Control_Mode = Float_Backup
    Alarm "Wet well level signal fault"
ELSE
    Level_Used = Level
    Last_Good_Level = Level

What to do when the level is bad

The correct fallback depends on the station, and it should be a documented decision in the control narrative rather than whatever the programmer chose that day.

1. 01Fall back to float control. If the station has start and stop floats, use them. The station runs crudely but it runs.
2. 02If only a high level float exists, use it to start pumps and run for a fixed timed period, then stop and wait for the float again.
3. 03Alarm immediately and distinctly. "Level signal fault" must not be buried among level alarms.
4. 04Never continue controlling on a value you have identified as invalid.
5. 05Never simply stop pumping and wait for someone to notice. That guarantees an overflow.

Variable speed level control

With drives, the station can hold a level rather than cycling between two. A PID loop modulates pump speed to keep the well near a target, adding the second pump when the first reaches maximum speed.

Two constraints matter. Minimum speed must stay above the point where the pump cannot overcome static head, or it spins without moving water. And a station that holds a constant level has constant detention rather than a full-to-empty cycle, so scour velocity in the force main has to be considered separately, often with a periodic forced high-speed run.

Reading a level trend

Wet well level trended against pump run status is the single most useful diagnostic display a utility has. The shape tells you what is wrong before anyone drives to the site.