How to Commission a Plunger-Lift Controller Cycle

Back to Blog

What You Need Before the First Automated Cycle

Commissioning starts with well data, not controller menus: the plunger type and its expected fall behavior, the tubing depth, and the operating guidance from production engineering for this well - because every timer you set encodes a physical claim about how long a plunger takes to fall through gas and liquid to bottom, and that claim is well-specific. A plunger that has not reached bottom when the well opens accomplishes nothing except wear; timers copied from the offset well up the road embed the wrong physics. Gather the wellhead's mechanical state too: lubricator condition, spring or bumper integrity, and catcher function, since the lubricator is what absorbs every arrival.

Then prove the instrumentation while the well is still under manual control. The arrival sensor - typically a magnetic or inductive sensor clamped at the lubricator - must register the plunger dependably: verify its mounting location and orientation per the sensor manufacturer, and exercise it, whether by test magnet, by manually cycling the plunger where practical, or by the sensor's own test function, until its response is trusted. A controller that cannot see arrivals cannot run min-max cycle logic at all; it will either strand the well closed or flow it blind. If the controller uses casing, tubing, or line pressures in its logic, confirm each transmitter reads sensibly against gauges before the automation gets a vote.

Set Conservative First Timers and the Non-Arrival Response

The first automated cycle should be deliberately unambitious. Set the close time long enough that the plunger has unquestionably reached bottom and pressure has built beyond the minimum the engineering guidance calls for; set the open period bounded by a minimum time - so a fast arrival still sweeps liquid and stabilizes flow - and a maximum time, so a slow or absent plunger cannot leave the well flowing indefinitely against the design. Where the controller supports pressure-based open and close criteria alongside timers, commission on the conservative side of the engineering numbers; the optimization pass comes later, after the cycle has proven itself mechanically.

Before walking away, configure the failure branch: the non-arrival response. When the open period expires without an arrival, the controller should close the well, extend the subsequent buildup, and - after a configured number of consecutive misses - stop cycling and raise an alarm rather than hammering a well that has told it something is wrong. A missing plunger, a stuck plunger, a failed sensor, and a loaded-up well all present as non-arrival, and the response that protects the well is a conservative close-and-alert, with the alarm actually landing somewhere a human sees it. At an unmanned site, that means verifying the non-arrival and arrival-count points reach the SCADA system - a platform such as Merobix trending arrivals, non-arrivals, and casing pressure per cycle gives the remote operator the same story the on-site commissioning tech watched in person.

Watch the First Cycles, Then Tune Gradually

Stay for complete cycles - plural. Log what the controller saw and what the well did: the pressure at open, the open-to-arrival time, the arrival's character at the lubricator, the flow period's behavior, and the pressure decline into the next close. Arrival timing is the central diagnostic: the controller derives an average rise velocity from tubing depth and open-to-arrival time, and both extremes are findings. Arrivals that come violently fast hammer the lubricator and spring and suggest the well opened with more stored energy than the lift required; arrivals that barely make it, or arrive late, suggest insufficient buildup or a liquid load near the limit of what the cycle can carry. Target arrival behavior is a site- and equipment-specific number that comes from the plunger and lubricator manufacturers' guidance and the site's operating practice, not from a universal table.

Tune the way wells reward: one variable at a time, in small steps, across enough cycles to see the response. The classic commissioning mistake is chasing production immediately - trimming close time cycle after cycle until the plunger starts missing bottom, then diagnosing the resulting non-arrivals as a controller fault. The buildup the close time buys is the energy budget for the next lift; spend it too thin and the cycle collapses. The other repeat offenders: ignoring non-arrival alarms because the well usually recovers, leaving the sensor's mounting unverified so intermittent missed detections masquerade as well problems, and treating a neighboring well's settings as portable. A commissioned cycle is one where consecutive arrivals are consistent, the non-arrival branch has been demonstrated, and the remote trend tells the same story the wellhead does.

Frequently Asked Questions