What Is a Float Switch?

Back to Blog

How a Float Switch Works

A float switch pairs a buoyant float with a switch mechanism. In a common magnetic design, the float contains a magnet and slides on a stem; as the float rises to the trip height, its magnet actuates a sealed reed switch inside the stem, changing the contact state. Other designs use a hinged float arm that mechanically toggles a microswitch, or a mercury-free tilt element in a tethered float. Because the actuation is mechanical and the switch contacts are sealed, float switches are rugged, need no external power to sense, and fail predictably.

A single float switch detects one threshold - level reached or not reached - and reports it as a discrete contact. Installations often stack several floats on one stem, or use several separate switches, to mark distinct setpoints such as pump-on, pump-off, high-high alarm, and low-low alarm. The contacts wire straight into a discrete input on a PLC or RTU, or directly into a motor control or shutdown circuit.

Where Float Switches Fit in Oil and Gas

Because they are simple and independent of any transmitter's electronics, float switches are widely used as safety and control point-level devices: high-level and high-high shutdown on tanks and vessels, low-level pump protection, and pump start-stop control on sumps and water tanks. They are frequently used as an independent backup to a continuous level transmitter so that a critical high-level trip does not depend on the same instrument that drives normal control.

Their discrete contact wires into a PLC or RTU digital input, or directly into an interlock. A cloud SCADA such as Merobix reads that switch state from the controller over Modbus or DNP3 as a digital status point, so an operator sees a high-level or pump-fault condition and can be alarmed remotely. The float switch does the sensing and, often, the hard-wired trip; SCADA provides the visibility and notification on top of it.

Selecting a Float Switch for the Service

Selection starts with the liquid, not the switch. The float must be buoyant in the actual process fluid, so its rated minimum specific gravity has to sit below the fluid's, with margin for temperature and composition swings; the float's figure is on the manufacturer's datasheet and the fluid's is site-specific. Materials come next - the float, stem, and any gaskets must tolerate the fluid chemistry and temperature - followed by the pressure and temperature rating of the process connection itself.

Electrical selection is the other half. The contact arrangement (normally open, normally closed, or changeover), the contact rating against the real circuit load, and the hazardous-area situation all matter. Reed contacts switching a low-level signal into a controller input are a different duty from contacts breaking an inductive pump-starter coil, and in classified areas the circuit design - intrinsically safe loops or suitably rated equipment - follows the site's area classification drawings and the applicable code, with the design signed off by qualified personnel.

Commissioning Checks That Catch Most Problems

1. Verify the float moves freely through its whole travel with no rubbing on the stem, chamber wall, or nearby internals.
2. Prove the trip by moving real liquid past the setpoint where the process allows, or by manipulating the float per the site's test procedure, and confirm the contact change at the controller input, not just at the switch.
3. Confirm the installed orientation matches the design - many designs are direction-sensitive, and a float assembled inverted trips backwards.
4. Check the as-found contact logic against the loop drawing: which state is normal and which is alarm.
5. Record the as-left trip level and contact configuration in the loop documentation.

The end-to-end proof in step two is the one most often skipped. A float that clicks convincingly at the switch can still be dead at the controller because of a wiring, polarity, or input-configuration fault, and the entire value of the device is that the trip works when demanded.

Failure Modes and Their Signatures

Float switches fail in a small number of characteristic ways, which makes a signature table useful during troubleshooting:

Failure mode	Typical signature
Float stuck from solids, wax, or scale buildup	No trip despite level clearly past the setpoint
Punctured or waterlogged float	Float rides low or sinks; trip late or never
Degraded or welded contact	Chattering input, or a contact frozen in one state
Process density change	Trip point shifted from the documented level
Wiring or input fault	Switch proves good locally but the controller never sees the change

Because a failed point-level switch is silent until demanded, periodic proof testing at an interval set by the site's procedures is the only way to know it still works. Where the switch protects against a hazardous overfill, the testing discipline and trip philosophy belong with the site's safety instrumented system practice rather than with informal maintenance.

Contact Sense and Fail-Safe Wiring

The safer convention for alarms and trips is a circuit that is closed and energized in the healthy state and opens on alarm. Wired that way, a broken conductor, a lost supply, or a corroded terminal produces an alarm rather than a hidden dead switch, converting the most common wiring failures into detectable events. Which physical contact achieves that depends on whether the float rests above or below its trip point in normal operation, which is why the loop drawing and the site's trip philosophy, not habit, should decide it.

Where a continuous reading also exists from a level transmitter, comparing the switch state against the transmitter's value is a cheap ongoing plausibility check between proof tests: a switch showing dry while the transmitter reads above the trip level is a discrepancy worth a work order, whichever device turns out to be wrong.

Frequently Asked Questions