Most small-to-mid operators could cut tank-gauging visits in half within 30 days. Same crew, same wells, half the touches. The lift does not come from new sensors. It comes from changing when a tank gets gauged from a fixed 3-day rotation to a fill-rate-driven trigger the operator already has the data to calculate.
The Cadence Is Wrong Before the Truck Leaves the Yard
The average small-to-mid operator gauges tanks on a 3-day rotation. Every tank gets the same visit cadence regardless of fill rate, regardless of well economics, regardless of whether a load is anywhere near ready. The 50-bbl heater treater on a 30-bopd well gets the same touch as the 400-bbl battery on a 200-bopd well. Most of those visits do not move a measurement.
Stand at the 50-bbl heater treater. The tank fills to gauge level once every six days. The route shows up every three. Half of those visits are a stamp on a clipboard, a glance at a sight glass, and a 20-minute drive back to the truck. Now stand at the 400-bbl battery on the 200-bopd well. The tank is ready for a load every 48 hours. The 3-day cadence misses every other load, oil sits long, and the working-capital line on the controller's balance sheet gets heavier than it needs to be.
The waste is not in the gauging. The waste is in the cadence. Gauging cadence built around a route schedule is a holdover from an era when no one could predict tank level overnight. That era ended when the SCADA, the meter, and the production-accounting system started producing enough signal to forecast fill within hours. The forecast has been available on most operator stacks for over a decade. Almost no small-to-mid operator gauges on it.
The Reframe: When, Not Which Day
Pump-by-priority does not change which tank gets gauged. It changes when. Every tank on the operator's footprint is scored overnight on three inputs: forecast fill level for the next 24 hours, measurement uncertainty since the last gauge, and downstream load readiness for crude trucking or pipeline scheduling. The score tells the crew which tanks need a gauge today and which can wait. The route is built around the scored list, not the calendar.
The pumper's day starts the same way as on a fixed-route operation. The crew opens the daily plan in the truck cab. The difference is what is on the plan. The 50-bbl heater treater on the 30-bopd well is not on it today. The 400-bbl battery is, because it crossed the 75% fill threshold overnight and the next load is scheduled before noon. The crew gauges the tank that needs gauging. The crew does not gauge the tank that does not.
This is the same operating logic that pump-by-priority applies to artificial-lift work, route optimization, and exception-based surveillance. Score every potential action overnight on cash-flow impact and risk. Rank the day around the score. Run the top of the list. The score function for tank gauging just has different inputs than the score function for a workover. Same engine, same morning artifact, different scoring layer.
Get the WorkSync Field Ops Brief
Monthly read for upstream + midstream operations leaders. Case studies, benchmarks, and what's changing in the field. Unsubscribe anytime.
Three Outcomes the Controller Can Monetize Separately
The lift from when-not-calendar gauging is not one number. It is three discrete outcomes, each of which a controller can model and fund independently.
Fewer Visits, Same Crew, Half the Miles
Gauging touches drop by roughly half on most operations inside 30 days. Same crew. Same wells. Half the miles to and from the tanks. The recovered hours land in two buckets: high-value field response (the 200-bopd well drifting 8% off forecast moves to the top of the day) and crew capacity (the asset that needed a second pumper for the 3-day rotation now runs on the existing headcount). The labor line on the operating budget either drops or, more commonly, gets reallocated to work that moves the production number.
Drive miles drop. Lone-worker hours on the road drop as a side effect. TRIR trends improve over quarters as the lowest-value, highest-mileage visits come out of the rotation.
Better Data, Faster Close
Measurements happen when the level is actually moving. The signal-to-noise ratio on every gauge improves because the gauge is taken at the level the controller cares about, not at whatever level happens to be in the tank on the third day of the route. Reconciliation between SCADA, the meter, and production accounting tightens. Lease accounting closes in days, not weeks.
This is where the CFO funds the pilot if the operating budget does not. A 7-day reconciliation lag costs an independent two things: trapped working capital and the optionality on hedge timing. Closing in two to three days releases both. On a 5,000-well asset the working-capital recovery alone is material to free cash flow inside the quarter.
Less Oil on the Books
Buffer capacity is a function of measurement uncertainty. The wider the uncertainty band on a tank's level, the more buffer the operator carries to avoid running over or running dry. Cut the uncertainty in half (which is what gauging-when-the-level-is-moving does) and the buffer comes down. The freed working capital is real. The trapped barrels are released back into the sales schedule.
This is the most direct line from tank-gauging cadence to the controller's balance sheet. Less buffer is less working capital. Less working capital is more free cash flow. The math is straightforward enough that a controller can run it on a single asset before signing the pilot.
What the Stack Actually Needs (and What It Does Not)
The most common objection inside the first conversation is the assumption that fill-rate gauging requires new sensors. It does not.
What the score function reads:
- SCADA tag-level data for tank level, run statuses, and meter readings
- Production accounting allocations at the tank-battery level
- Crude trucking or pipeline scheduling data for load readiness
- Field-data-capture entries from the most recent gauges, including the pumper's voice-captured notes
- Equipment-history data from the EAM or CMMS for tank-side maintenance flags
What the score function does not require:
- New radar gauges
- New level-transmitter installs
- A SCADA upgrade
- A historian migration
- New headcount
The Work Engine reads from AVEVA PI, Ignition, Cygnet, eLynx, and 15-year-old historian installs without changing the historian configuration. Operators on legacy SCADA stacks have deployed the scoring loop in a week. The bottleneck is whether the data is reaching the score function, not whether the level transmitter is current.
What Changes in the Operations Center
Three things change in the morning meeting once the score is running.
The superintendent stops rebuilding the gauging route in a spreadsheet. The Work Engine publishes the ranked tank list by 6 AM. The superintendent adjudicates the top items, flags any tank that the score over-ranked or under-ranked, and feeds the correction back into the engine. The score gets sharper inside the first week as it absorbs operator corrections.
The asset manager stops adjudicating "did the pumper hit every tank on the route" and starts adjudicating "did the crew gauge the tanks that needed gauging." Different question, different morning artifact, different management cadence. The asset manager's job gets easier because the question is now answerable.
The controller gets a daily reconciliation status that closes against the previous day's measured gauges, not against the previous day's scheduled visits. Lease accounting follows what actually happened in the field. The close cycle compresses.
What the Pumper Sees in the Truck Cab
The pumper opens the route on a tablet or phone. The route is no longer a fixed sequence of tanks in calendar order. The route is the top of the day's tank score, ordered for the crew, the basin, and the truck. A 50-bbl heater treater that has not moved a measurement in five days is not on the route. A 400-bbl battery that crossed 75% overnight and has a 10 AM trucking pickup is at the top.
Voice capture is the second visible change. The pumper speaks the observation while standing at the tank. "Top gauge 7-foot-2, no water cut change, bottom valve dry, sight glass clear." The observation flows into the score before the next morning's plan. The handwritten ticket, the radio call to the office, the gut sense about a tank that "felt wrong yesterday" all become scoreable inputs.
The cultural friction usually shows up around the second week. Pumpers who built routes around personal knowledge of the field will sometimes resist a ranked plan that skips a tank they "know is fine." The pattern that resolves it is letting the score show the math. When the 400-bbl battery the pumper gauged on the 3-day rotation actually had a backside leak that the new cadence caught 36 hours earlier, the pumper becomes the most vocal advocate for the score.
The 30-Day Shape
The shape of a tank-gauging pump-by-priority pilot is the same shape as the broader 4-Week Pump-by-Priority Pilot, with the score function tuned to tank-side signal.
- Week 0. Pick the controller-signed metric. Most operators sign on gauging visits per week, lease-accounting close cycle, or working-capital recovery from inventory drawdown.
- Week 1. Read-only integration onto SCADA, production accounting, trucking schedule, and EAM. The score function runs offline against the trailing 90 days so the operations team can see what the ranked plan would have looked like.
- Weeks 2 to 3. Live ranked plan in the truck cab by 6 AM. Voice capture flowing back into the score. Operations-center adjudication tightening the rank.
- Week 4. Measure against the Week-0 baseline. If the metric moved, the operator signs the annual. If it did not, the operator walks away.
No license fee in the 4 weeks. The Impact Guarantee removes the bias to continue. The pilot is the cleanest test of the approach on the operator's tanks, basin, and crew.
What This Is Not
It is not a SCADA refresh. It is not a level-transmitter install. It is not an AVEVA migration. It is not a P2 or Enertia replacement. It is not a multi-quarter system implementation.
It is also not a generic AI deployment. The lift in the first 30 days comes from changing when the crew gauges. The score function is the artifact that makes "when" a defensible answer instead of a guess. The model layer compounds the lift over the rest of the year by getting sharper at predicting fill, but the first 30 days do not need the model to be clever. They only need the cadence to come off the calendar.
What the Operator Owns at Day 28
By the end of the four weeks the operator owns:
- A baseline of the chosen tank-gauging metric and a measured delta over 28 days
- A daily ranked tank list published to every truck cab by 6 AM
- A score function tuned to the operator's tanks, basin, well mix, and trucking cadence
- Voice-first field capture flowing back into the score
- An adjudication workflow in the operations center
- A reconciliation feed that closes lease accounting against measured gauges, not scheduled visits
- A one-page operating-change document the controller signed
No new sensors. No SCADA upgrade. No new headcount. The bottleneck was never collection. It was scoring. The pilot moved the scoring layer onto the data the operator already had.
This is the shape of the deployed reference at a top 25 private producer running over 5,000 wells across the Western Anadarko, Permian, and Wyoming. The metric the controller signed for in Week 0 was working-capital recovery on the tank-inventory line. The metric moved. The annual signed.
Where to Start
The shortest path is one conversation. Send your asset count, your basin, and your tank count. WorkSync sends back a per-asset gauging-math estimate, a draft Week-Zero metric, and a one-page pilot scope inside 30 minutes.
Request Your Free Trial. 4-week adoption pilot on the stack you already own. The metric is signed on Day Zero. The exit is clean if the metric does not move. The operator owns the integration either way.
For operators who want the personalized gauging math before requesting a trial, the Adoption Math intake returns the per-asset estimate without committing to the pilot scope.
FAQ
How does pump-by-priority cut tank-gauging visits in half without new sensors? By changing when a tank gets gauged from a fixed calendar cadence to a fill-rate-driven trigger. The score function forecasts each tank's level for the next 24 hours using the SCADA, meter, and production-accounting data the operator already has. The crew gauges the tanks where the measurement is actually moving. Tanks that have not changed level since the last gauge come off the route. The lift is in the cadence, not in the hardware.
What if our SCADA does not have radar-level gauges on every tank? It is not a blocker. The score function reads any combination of level, meter, and allocation data the operator's stack already produces. Tanks without continuous level signal are scored on meter throughput and last-gauge measurement. The accuracy of the rank improves with sensor density but the operating change is durable on legacy stacks.
How does this affect lease accounting and the close cycle? The close cycle compresses because measurements happen when the level is moving and reconciliation tightens. Operators currently running a 7-day reconciliation lag typically close in two to three days inside the pilot. The math goes to the controller's balance sheet through working-capital recovery and faster hedge-timing optionality.
Will pumpers accept skipping tanks on the route? The pattern is consistent. Initial resistance in Week 2 resolves once the score catches something the calendar cadence would have missed. A tank flagged on fill-rate criteria that turns out to have a backside leak or a load-readiness signal the calendar route would have skipped is the moment the pumper becomes an advocate. Voice capture keeps the pumper's field knowledge inside the score, so the crew is not bypassed, just better directed.
Can the pilot run on a single battery or route before expanding? Yes. A typical pilot covers one route of 80 to 150 wells and 20 to 60 tanks across one or two pumpers. The score function trains on the full asset but the operating change is contained to the pilot route until Day 28. Expansion to the full asset is a 60-to-90-day rollout after the annual is signed.
What metric should the controller sign for in Week 0? The most common starting metrics are gauging visits per week on the pilot route, lease-accounting close-cycle days, and working-capital recovery from tank-inventory drawdown. All three are easy to instrument off existing SCADA and production accounting, hard to fake, and directly tied to free cash flow. Controllers consistently sign on at least one.
Sources
- Alvarez & Marsal (2015). "The Advantages of Exception-Based Surveillance." Clevenger, Shannonhouse, Patterson. The foundational industry whitepaper on cadence-driven versus exception-driven field operations.
- ExxonMobil and SLB. SPE Artificial Lift Conference, 2024. Permian closed-loop deployment, 1,300+ wells, 2% uplift, no equipment changes.
- ConocoPhillips. JPT, 2024. Montney IOCaaS case study, 3 to 4% above forecast, trending to 6%.
- Chevron. JPT, 2024. Kaybob Duvernay IOCaaS case study, 5% LOE reduction year one.
