A rod-pumped well produces well when the downhole pump fills completely with fluid on every stroke. When the reservoir can no longer feed the pump fast enough, the barrel only partly fills, the plunger drops onto liquid or gas with a jarring impact, and the equipment starts to punish itself. The pump-off setpoint is the number the pump-off controller watches to catch exactly that condition. This guide explains what the setpoint measures, how an operator picks its value, and how it protects both the pump and the rod string from pumping a well dry.
Pump-Off Setpoint in one line: A pump-off setpoint is a threshold value - most often a fillage percentage read off the dynamometer card, sometimes a load value - that a pump-off controller compares against every stroke. When measured fillage falls below the setpoint for a set number of strokes, the controller concludes the pump is no longer filling with liquid and shuts the well in so the fluid level can recover.
The dynamometer card plots rod load against position over one complete pump stroke, and its shape is a fingerprint of what the downhole pump is doing. When the pump fills completely, the card is full and boxy, with the traveling valve carrying load through most of the upstroke. When the barrel only partly fills, the card develops a rounded lower-right corner: the plunger travels through empty or gassy space before it contacts liquid, and load transfer happens late. A pump-off controller processes that card, calculates a fillage percentage, and compares it to the pump-off setpoint. A fillage setpoint of, say, seventy percent means the controller expects the pump to be at least seventy percent full and treats anything below that as an incomplete fill.
Some controllers use a load-based setpoint instead of, or alongside, fillage. Because a partly filled pump changes where and how load appears on the card, a load threshold at a chosen card position can flag the same condition. Modern controllers that compute a downhole card from surface measurements can locate the fluid-load transfer point directly and derive fillage from it, which is generally more reliable than reading surface load alone. Whatever the underlying signal, the setpoint is a single line the controller draws: on one side the pump is fed, on the other it is starving.
Picking a pump-off setpoint is a balance between producing every barrel the well will give and not beating the equipment. Set it too low - permit very poor fillage before shutting in - and the pump pounds fluid stroke after stroke, wearing the plunger and barrel and fatiguing the rods. Set it too high - demand nearly full fillage - and the controller shuts the well in prematurely, giving away production the reservoir could still deliver. Operators typically start from a value in the range where fillage clearly indicates the pump is outrunning inflow, then tune it against the well's own behavior over several days, watching how fillage trends across each pumping cycle.
The setpoint rarely acts on a single stroke. A well's fillage naturally jitters from stroke to stroke because of gas breakout, slippage, and normal variation, so a controller waits for the condition to persist. A stroke-count or timer requirement - fillage must stay below setpoint for a number of consecutive strokes, or for a set time - filters out the noise so a momentary dip does not trip a shutdown. Together the fillage setpoint and the stroke count define the pump-off event: how starved the pump has to look, and for how long, before the controller decides the well has genuinely pumped off and stops the unit.
When fillage drops below the pump-off setpoint for the required strokes, the controller shuts the unit down and starts an idle timer. During that idle period fluid from the reservoir accumulates in the annulus, the working fluid level rises, and the pump is given a fresh charge of liquid to lift. The controller then restarts the unit, and on the next cycle the pump fills better because it has fluid to pull. This shut-in and restart loop is what pump-off control exists to do, and the setpoint is the trigger that decides when each cycle ends. Without it, the pump would either run continuously and pound fluid, or an operator would have to guess at a fixed on/off timer that ignores what the well is actually doing.
In a modern field these events do not stay hidden at the wellhead. The controller reports each shut-in, restart, and the fillage that triggered it up through the field RTU to a SCADA host, and with a cloud SCADA platform such as Merobix that stream lands where an engineer can see every well at once. Rather than driving to each location to read a local display, the team watches fillage trends, pump-off counts, and runtime percentages across the whole field from a browser, and can spot a well whose setpoint is mistuned - shutting in too often, or clearly pounding without ever tripping. Surfacing the setpoint and the events it produces this way turns a local protective threshold into a fleet-wide optimization signal.
There is no single correct value - it depends on the well's inflow, the pump, and how the operator weighs production against equipment wear. Many operators begin somewhere in the range where fillage clearly shows the pump outrunning the reservoir, then tune the number against the specific well over several days. The goal is to shut in before the pump pounds fluid hard, without shutting in so early that recoverable production is left behind.
A well's fillage naturally varies from stroke to stroke because of gas breakout, slippage, and normal fluid dynamics, so a single low reading does not reliably mean the pump has pumped off. Requiring fillage to stay below the setpoint for a number of consecutive strokes, or for a set time, filters out that noise. This prevents nuisance shutdowns from momentary dips while still catching a genuine, sustained pump-off condition.
A fillage setpoint compares a calculated pump-fill percentage against a threshold, while a load setpoint watches rod load at a chosen point on the dynamometer card. Both aim to detect the same physical condition - an incompletely filled pump - but fillage derived from a computed downhole card is generally the more direct and reliable indicator. Some controllers use both together for confidence before shutting the well in.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.