Dynamometer Card Signature Screen
Read a sucker-rod pump card: fillage on the plunger stroke, card geometry, and a ranked shortlist of the twenty named failure-mode signatures
Card Input
Reference signatures are synthetic cards drawn at the centre of each condition's parameter range.
Twenty named pump conditions.
0 draws the canonical card for that condition. Any other whole number draws a different card from the same condition's parameter range — useful for seeing how much a signature moves.
Card Measurements
Signature shortlist
What this screen does not do
- It does not diagnose a pump. It ranks a card's geometry against reference signatures. A card can sit closest to a signature and the pump still be doing something else.
- It carries no accuracy figure, deliberately. The reference library is synthetic. Checking generated cards against their own generators measures self-consistency, not field performance, and there is no benchmark against labelled real cards behind this tool to quote.
- It assumes one fault at a time. A gassy well that is also pounding does not match any single reference signature.
- It cannot separate every mode. A paraffin restriction with shallow dents, and a travelling-valve closure delayed over only the very bottom of the upstroke, both sit on an otherwise normal card body and can read as normal. Absence of a signature is not evidence of a healthy pump.
- Stroke-length conditions need real units. Normal, tubing movement and plunger undertravel differ only in absolute plunger travel. A card pasted in normalised (0–1) units cannot be screened for them.
- Confirmation is field work. A fluid shot, an intake-pressure reading, the runtime recorded with the card and a valve check are what turn a shortlist into a diagnosis.
About This Screen
A dynamometer card is a closed loop of load against position over one pump stroke. On the pump card its four corners correspond to the four events of the pump cycle: the standing valve closing and the fluid load transferring onto the travelling valve at the bottom of the stroke, the plunger lifting the column, the travelling valve opening and the load transferring back at the top, and the plunger falling through the fluid to the bottom again. Pump conditions distort that loop in characteristic ways, and this screen measures the distortion.
Pump fillage — on the plunger stroke
Fillage is net plunger travel divided by gross plunger travel, both taken from the pump card. Net travel is the position of the bottom-right corner above the bottom of the stroke; that corner marks where the travelling valve finishes opening, and on a poorly filled pump it sits well below maximum position. It is located as the point on the downstroke branch that maximises (retained stroke − remaining load) in normalised coordinates.
Fillage is never computed on the surface stroke. Rod stretch under the fluid load and rod-string overtravel put the polished rod and the plunger feet apart on a deep well, so a surface-stroke fillage is wrong by the rod-string deflection — and it is wrong in the unhelpful direction, flattering a starved pump.
The measurements the ranking uses
- Corner clearances — distance from each bounding-box corner to the nearest point on the card, as a fraction of the normalised diagonal. A healthy card clears its corners slightly because fluid-load transfer is not instantaneous; a corner that is gone or one the card actually reaches is the signal.
- Fill ratio — enclosed card area over its bounding box. A collapsed or gutted card does not fill its box.
- Branch slopes — least-squares slope of normalised load against normalised position over the middle of the upstroke and the downstroke, with the turnarounds excluded so the slope measures the branch rather than the corner rounding.
- Downstroke transfer width — the fraction of stroke over which the downstroke sheds its fluid load. This is the one measurement that separates fluid pound (abrupt) from gas interference (gradual).
- Load range against mean load — absolute scale. A parted rod string draws an ordinary-looking loop at a small fraction of working load.
- Plunger stroke length — in the card's own units, because tubing movement and plunger undertravel are otherwise identical to a normal card.
- Peak and minimum load prominence, and where on the stroke they fall — an impact spike is a distinct peak at one end of the stroke, not a raised plateau.
- Trace roughness and periodicity — roughness alone cannot separate sand abrasion from unit vibration. Vibration is a wave and autocorrelates; sand is grain-by-grain scatter and does not.
Each measurement is standardised across the reference set, weighted, and the card is ranked by weighted Euclidean distance to each signature. A mode carrying more than one reference variant — a bent barrel binds at either end of the stroke and so draws two mirror-image cards — is ranked by its closest variant. The shortlist names, for every signature, the three measurements that pulled the card towards it and the three that pushed it away, so the match can be checked by eye rather than taken on faith.
Reference signatures
The reference cards are synthetic and drawn from published signature descriptions, not from any client's wells. They exist so the geometry a screen looks for is written down and inspectable. See the worked examples for all twenty drawn side by side with the measurements that identify each one, and the method page for the wave equation, the fillage and pressure relationships, and the full limits.
Need Rod-Lift Optimisation Support?
For fleet-scale card screening, surface-to-pump conversion against your own rod-string data, and pump-off controller review, we work from your recorded cards and hand back the pipeline.