Which wells to work over, which to leave, and which to let go.
Well-by-well production engineering for fields where a workover has to pay for itself. Send one dynamometer card file and get back a real diagnosis — including which parts of the card are not trustworthy enough to diagnose on.
This list did not come from a market study. It came from reading what operators running shallow, conventional wells wrote about their own fields — the cards they pulled, the bearings they lost, the parts they could not find.
We work rod lift and production engineering as software: physics that runs on your well data, not a dashboard subscription. Where we can help, this page says so plainly. Where we cannot, it says that too.
Three things we do well
Read the card properly
Full surface-to-downhole conversion, fillage computed on plunger stroke rather than surface stroke, failure mode with a confidence figure — and an explicit statement of which segments of the derived card carry enough numerical confidence to diagnose on.
Rank the field
Nobody holds 150 wellbores in their head, so the marginal ones stay invisible until they fail loudly. We score every well you send and return one ordered list: which to drive to this week, and why. No new hardware on the lease.
Price the decision
Keep, work over, shut in, or plug — scored on one basis, with the plugging liability carried in every case. Leave P&A out of the sum and every answer quietly tilts toward "keep pumping".
What we keep hearing
- The well pumps off in five minutes and there is no controller on it. "Set a shutdown below 65–75% fillage" assumes hardware the well does not have, and is not tuned to your pump, rod string or inflow. The right setpoint is a calculation.
- Is that noise on the card real, or is it the instrument? On an accelerometer-based dyno, position is integrated twice from acceleration, so error accumulates across a stroke. Some of what looks like a pump tag is arithmetic.
- The low-volume wells get no attention until a neighbour calls. Not an equipment problem — a ranking problem.
- The walking beam left and took the saddle bearing with it. Torque and gearbox loading against the unit's actual rating finds the units running outside their envelope before the structure tells you.
- Repeated tubing splits, never the same joint. Fluid pound puts the lower rods into compression on the downstroke; buckled rods bear on tubing. If that holds, the pump-off fix is also the tubing fix, and rotators are treating a symptom.
- One missing seat nipple idles a rig day. A rod-lift workover bill of materials is something every operator in a basin holds a partial copy of.
- "Workovers will add 30 BOPD" — will they? A capital partner asks which wellbores, how much, and what the error bars are. That is inflow and field-health work, and it is the number that a flat claim cannot defend.
Worked example — the plugging bill is deferred, not avoided
A marginal well is not competing against zero. It is competing against paying the plugging bill today, and deferral has real time value. That is why a well can lose money every month and still be worth holding — and why "shut in" is sometimes right when both producing and plugging are wrong.
| Well | Month-1 operating cash | Breakeven oil price | Way forward |
|---|---|---|---|
| 8 BOPD, 30 BWPD | +$9,412 | $18.38/bbl | Keep |
| 4 BOPD, 60 BWPD | +$3,154 | $33.76/bbl | Keep — 127 months left |
| 1.5 BOPD, 90 BWPD | −$1,014 | $91.79/bbl | Shut in |
| 0.6 BOPD, 120 BWPD | −$2,953 | $260.03/bbl | Shut in |
Illustrative cost deck — $65 oil, $45k plugging, 12.5% royalty, $0.75/bbl water, 8%/yr decline, 10% discount. Not any real operator's numbers. Note the bottom two: producing them destroys value, yet shutting in still beats plugging by roughly $16,000 each, purely from deferring the plugging bill.
An example of the kind of thing we will tell you. Preparing a card review this month we found a defect in our own fillage calculation: past a certain pump-off severity it reported a pounded-off pump as full. Severe, fast pump-off is exactly the regime a controller setpoint exists to catch, and it failed toward "healthy".
It is fixed, with a regression test, and the four vendor-analysed reference wells still agree to within 0.85 fillage points. We mention it because a firm that never reports finding anything wrong with its own tools is not checking.
How to start
Send one well's card file — whatever your dyno or handheld exports. No install, no telemetry, no account.
You get back the downhole card, fillage stroke by stroke, the diagnosis with a confidence figure, and the parts of the card we do not trust. If you also send a full-pump card from the same well, we calibrate the fillage scale first — a "shut down at 75%" rule means something different against a scale that does not top out at 100%.
If it is useful, send the field. If it is not, you have lost an email and we have learned something about your wells.
What we do not do: find you money, or advise you on your bonding or plugging obligations. We are engineers, not a placement agent, and the obligation is a question for your counsel. We model the cost.