botminds.aiRequest a demo
WorkflowProduction Operations & Oil Sands

Production Deferment Management

Production deferment software that traces each lost barrel from the shift's entry to its root event, measured against what the wells could really deliver.

See the demoBook a demo
Follow forty wells back to one compressor trip ↓
The problem

Production deferment booked well by well makes one compressor look like forty

When a lift gas compressor trips, the night shift closes in every well on it. By the end of a twelve-hour shift the supervisor is bringing wells back, and the losses go in one entry per well, under whatever code fits, at plan rates and round hours.

The month report then ranks the wells as the problem and the compressor barely shows. Barrels are overstated, a planned job that did not start still counts as planned, and the ledger drifts away from production accounting. The cause sits across four systems, and nobody has the hours to line them up.

Built forProduction engineersOperations managersShift supervisorsProduction accountants

The Problem Today

One compressor trip that shuts in forty wells is booked as forty well problems, and nobody ties it back to its cause.

$47.5M
a year for one point of deferment at 186,000 barrels a day our estimate
24%
of one offshore basin's potential not produced in 2025 published

Tied together by hand today

  • Shift logs
  • Trips
  • Work orders
  • Well tests
  • Measured against plan ratesBarrels from plan rates and round hours, not tests
  • Planned work that overranOverruns and jobs not started still count planned

The Solution

  1. Each entry read, unchangedEach loss entry and code read as typed; the original stays as it was
  2. The records gatheredShift log, trip record, work order and well tests, each line cited
  3. Measured against testsEach well's accepted test times its own hours down, not a plan rate
  4. One cause, two signaturesForty entries become one event; engineer signs, manager countersigns

Built-in controls

  • Shift entries are never changed or deleted
  • Nobody countersigns their own recode
  • The month posts only when it ties

What You Get

Fix money goes to the real cause, and the month ties to production accounting before it posts.

  • Each link citedEach value opens the log line or record it came from
  • Fixes on the real causeOpen fixes ranked by the barrels behind them
  • Plant questions reach a personQuestions on running plant reach a named person
The morning after a trip

Production deferment from forty shift entries to one signed root event

The lift gas compressor on one platform tripped in the night and took forty gas-lifted wells with it. The day shift booked the losses at shift end. Next morning, from 08:10, the production engineer and the operations manager work it through.

demoThe field, its platforms, wells and people, and every barrel count are sample data.

  1. 0108:10

    40 well problems under 11 codes, and not one names a compressor

    Production engineer

    Each bar is one well's lost hours as the shift booked them, grouped by the code the supervisor chose: low rate, liquid loading, lift valve, slugging, choke trouble and six more. Together they book 9,536 barrels. Read this way, the field had a bad day at forty wells, and the month report will rank wells as the top unplanned cause.

  2. 0208:12

    One trip at 03:12, down 10 h 42 min, and 8,210 barrels in the event

    Production engineer

    She runs the agents on the night. They read the entries, the shift log, the trip record, the work order and the well tests, and the bars collapse into one block under the outage, each shrunk to the time that well was really down. One well's late tail peels off as its own lift valve fault. The proposal is 96% sure.

  3. 0308:12

    The trip record names the first out: packing vent high high, throw 3

    Production engineer

    Four records must be opened before anything can be signed. In the control system's trip record: the packing vent alarm, the trip 18 seconds later, the unit valves closing and lift gas to the platform header falling from 19.7 MMscf/d to zero. Restarted at 13:15, the compressor was loaded at 13:54.

  4. 0408:12

    Rod scored over the packing travel, so new packing alone would not hold

    Production engineer

    The work order says what failed and what was done: seal rings broken, rod scored, piston rod and packing case replaced from spares, rod run-out checked. Four mechanics, 7 h 30 min on the job after isolation and purge. It is the third packing failure on this throw this year, and the follow-up names crosshead alignment and packing lube.

  5. 0508:14

    A wireline job booked as planned did not start; 199 barrels move

    Production engineer

    One of the forty was entered as planned wireline, 239 barrels for 13.5 hours. The crew stood down at 08:00 and did not rig up, so the well was lost to the trip like the others. Measured from its accepted test and its own time down, it could have delivered 199 barrels. The supervisor's entry stays exactly as she typed it.

  6. 0608:18

    Three wells ruled before she can sign: one moved, one stale test, one tail

    Production engineer

    She moves the wireline well to unplanned. A second well's last accepted test is 79 days old against a 60-day limit, so she rules the seven-day rate of 460 over the old 595, which would have added 63 barrels. The third did not unload until 21:06; its 143-barrel tail stays with that well. The sign button checks each step.

  7. 0708:20

    Her signature is on it, and his countersign waits for his own opens

    Production operations manager

    She signs with a reason: same hours, same header, trip and work order agree. On his screen the panel shows her signature and two checks still open: every source opened by him, 0 of 4, and her three ruled wells opened by him, 0 of 3. Her opens do not count for him, and he cannot countersign his own work.

  8. 0808:22

    Signed by two people, and the ledger drops 1,183 barrels

    Production operations manager

    He opens the shift log, the trip record, the work order and the well tests, then the three wells, and countersigns. The ledger falls by the 1,183 barrels the shift booked too high, to 472,960. The forty entries now point to one root event, and each still reads as the supervisor typed it.

  9. 0908:24

    The month now ties to production accounting at 0.29%

    Production operations manager

    Potential for the month less allocated production says the wells did not produce 471,600 barrels. Against that, the ledger now holds 472,960: 0.29% apart, inside the 0.50% limit. Four close gates are still shut: other root events, the small recode lot, open questions and stale tests. Once they pass, the engineer prepares the close and he posts it.

What the forty wells could really deliver against what was booked

Each well's loss is its accepted test rate times its own time down, divided by 24 hours.

Booked by the shift, at plan rates and round hours9,536 barrels
Booked too high−1,183 barrels
Could really deliver, from tests and time down8,353 barrels
Of which in the compressor event8,210 barrels
Of which stays with one well's own lift valve fault143 barrels

demoSample data, from the working solution.

The month close before and after one countersigned root event

Only signed decisions move barrels. One signature on the compressor event brings the ledger inside the tie limit.

LineBeforeAfter
Potential, 186,000 barrels a day for 31 days5,766,0005,766,000
Less allocated production5,294,4005,294,400
What the wells did not produce471,600471,600
Deferment ledger474,143472,960
Apart, against a 0.50% limit0.54%0.29%

demoSample data, from the working solution.

Around the month

Deferment reporting for the month, by cause, entry and well test

What the engineer, the manager and the production accountant read once the night's event is signed.

Causes as entered and as signedRanked as the shifts entered them, wells lead at 140,463 barrels. As signed so far, compression leads at 84,640. Open proposals move nothing.
Loss entries, unchangedEach entry as the shift typed it, its code and barrels, the root event it traced to and where it stands. Search by well, code or words.
What the wells could deliverTested potential by platform, the tests out of date with their seven-day rates, and booked against tested for the widest gaps in the event.
Planned work against what happenedEach planned job against what happened: on plan, inside its allowance, overran by 11 hours, did not start. Barrels leave planned only when two people sign.
Top causes and open fixesInside compression, each event with its barrels, and the fixes still open with an owner, a due date and the barrels behind them.
Audit trailEvery read, proposal, ruling, signature and ledger move, with who did it and when. Original entries stay unchanged.

demoSample data, from the working solution.

What stays with the engineer, the manager and the shift

Agents read, measure and propose a cause. These decisions are people's.

A supervisor's entry is never rewrittenThe recode sits beside it and points to the root event. The supervisor can object to a merge of her entries.
No merge without a personThe root event tracer proposes. The production engineer signs, and the operations manager countersigns after opening the evidence himself.
No cause stated until it is provedA proposal under the confidence line reads not proved, and a question goes to the person who can settle it.
No stale test used without a rulingA test past the age limit needs a person to choose the seven-day rate or the old test.
Nothing posts to accounting by itselfThe close reconciler ties the ledger and lists the gates. The engineer prepares, the manager posts.

Four records behind one root event, and three more for the month

All are read only. Each link in a proposal opens the line it came from.

Production reporting systemLoss entries and codes, as typed
Shift logWells closing in and coming back, by time
Control system historianTrip records and first outs
Maintenance systemWork orders: problem, found, done, follow-up
Well test databaseEach well's latest accepted test and its age
Planning systemPlanned jobs with their hours and barrels
Production accounting systemAllocated production for the month close

What production engineers ask about production deferment

What is production deferment?

Production deferment is the oil a field could have produced and did not, split into planned, unplanned and external. This solution traces each booked loss to its root event, measured against each well's accepted test, and two people sign the cause before the month closes.

How is a well's loss measured?

From its latest accepted well test inside the age limit, times the time that well was really down, divided by twenty-four hours. A plan rate is not the basis, and a stale test needs a person's ruling.

Does it change what the shift supervisor entered?

No. Each entry stays as it was typed. The recode sits beside it and points to the root event, and the supervisor can object to a merge of her entries.

How does this fit with production loss accounting at month end?

The close reconciler ties the deferment ledger to potential less allocated production and shows the difference against the tie limit. The month posts only when five gates pass, prepared by the engineer and posted by the operations manager.

Is this production loss tracking or root cause analysis?

Both, for the losses already booked. It reads the downtime and deferment entries, groups those that share a cause, and shows what supports the link and what does not. Open fixes are ranked by the barrels behind them.

What happens when the agent is not sure?

Below the line the proposal reads not proved and cannot be stated as a cause. A question goes to the shift supervisor or the maintenance planner, and their answer moves the confidence.

Production Operations & Oil Sands

Other solutions for production and well performance

All 8 solutions →
Simultaneous Operations Planning

Conflicts the combined list cannot show, found in the job plans and settled before the day.

1-min demoSafety System Bypass Management

Every bypass on the control system has a key and a named person watching it, checked every shift.

1-min demoOperating Procedure Management

Procedures that match the plant, and a yearly signature that says what was read.

1-min demo

Bring one bad night
and last month's codes.

We will trace your booked losses back to their root events and measure them against your well tests.

Book a demoBrowse solutions