Non-productive time is the lost time everyone sees: it has a code, a line on the daily report and a cost. Invisible lost time hides inside operations that look productive, a connection two minutes slower than the best crew's or a trip stand that waits in the slips. This article works through both on a campaign of nine horizontal wells: the NPT Pareto, what the equipment share says about maintenance, and the connections and trips measured stand by stand against the best composite.
NPT, flat time and invisible lost time
Rig time splits three ways. Productive time advances the well as planned: drilling, and the connections, trips and circulation the plan needs. Non-productive time (NPT) is time lost to an unplanned event that stops progress: a failed tool, a rig repair, waiting on equipment or weather, losses, a stuck string. Invisible lost time (ILT) is the time a planned operation takes beyond what the same operation has been shown to take: the extra minutes in each connection, each trip stand, each BOP test.
NPT is found by coding. Someone writes the event on the daily report and gives it a code, and the totals are only as good as the codes. Two rigs that call the same top-drive fault "rig repair" and "equipment" produce two Paretos that cannot be added. ILT is found by measurement. Nobody writes "slow connection" on a report, so the time has to be measured from the rig's sensors, operation by operation, and compared with a target.
The two need different fixes. Most NPT comes from failures and hole problems, fixed by engineering, maintenance and planning. Most ILT comes from habits and practices, fixed by the crews once they can see it. Both need the same base: a rig-state track that says what the rig was doing every few seconds.
Measuring flat time from the sensors
A rig-state detector reads the surface channels (hook load, block position, bit and hole depth, rotary speed, flow, weight on bit and pressure) and classifies every sample: rotary drilling, sliding, connection, tripping in or out, reaming, circulating, static and so on. Strings of states then become operations. A drilling connection is measured bottom to bottom in three parts:
tconn = tw→s + tslips + ts→w
ILT = Σi max(0, ti − ttarget)
- tw→s
- weight to slips: from the last drilling sample to the slips set; reaming, circulating or a survey before the connection count here
- tslips
- in slips: the string hanging in the slips while the stand is made up
- ts→w
- slips to weight: from the slips pulled to drilling again, including any wait on a survey
- ttarget
- the target for that operation in that hole size: here the best offset well's average
Trip stands split the same way into pipe moving and in slips. ILT is summed only over time above target; an operation faster than target saves nothing in the sum. Implausible events, such as more than 20 minutes in slips on a drilling connection, are flagged as outliers and left out, because they are usually NPT or flat-time work that was never coded.
The target matters most. The best composite takes the best demonstrated value of each part, from whichever well achieved it: the weight-to-slips of one crew, the in-slips time of another. It is a connection nobody has made in one piece, built entirely from parts somebody has.
The example campaign
The example is a campaign of nine similar horizontal wells in North Oman, three rigs drilling three wells each. Every well has the same 6⅛″ lateral in a carbonate reservoir: 986 m from the 7″ liner shoe at 1,760 m to 2,746 m MD, 34 stands of 29 m. Each rig trips in through 1,740 m of casing, drills the lateral with a motor, circulates two bottoms-up at TD and trips out. The rigs differ in how they make a connection and how fast they trip.
| Rig 1 | Picks up and sets the slips; 2.75 min in slips on its first well; drills on as soon as the slips are out. Trip stands: 75 s moving, 55 s in slips |
|---|---|
| Rig 2 | Reams 9 m up and back down before every connection, 6 min; 2.0 min in slips; waits 1.5 min for the survey with the pumps on. Trip stands: 65 s and 40 s |
| Rig 3 | Circulates 2 min off bottom before every connection; 3.9 min in slips. Trip stands: 85 s and 80 s |
| All rigs | Second and third wells 6 % and 11 % quicker in slips and on trips; one stand in five slid |
Each lateral's 5-second surface log went through DrillIQ's KPI & Rig State engine, the same detector that runs on live data. It found 33 complete drilling connections and 152 trip stands in every well. The NPT rows from the daily reports were laid over the rig-state track with their codes, as a crew would enter them.
Where the NPT went
The campaign's 96.5 h of NPT, coded on the daily reports with DrillIQ's NPT codes and laid over each well's rig-state track (KPI & Rig State, final track). The top two codes hold 75 % of it.
Show the data
| NPT | |
|---|---|
| Downhole tool failure (3005) | 45.5 |
| Rig repair (3004) | 26.5 |
| Hole problems (3000) | 12.0 |
| Waiting on equipment (3001) | 8.0 |
| Weather (3003) | 2.5 |
| Waiting on third party (3002) | 2.0 |
The nine laterals took 759.5 hours, 562.4 of them drilling. NPT was 96.5 hours, 12.7 %. Two codes hold three quarters of it. Three downhole tool failures, two MWD tools and one motor, cost 45.5 hours, each one a round trip to change the tool, 14.5 to 16 hours apiece. Rig repairs cost 26.5 hours over seven events. Hole problems, two losses in fractured zones and one tight hole, cost 12 hours.
A Pareto by code tells you where to look; a Pareto by cause tells you what to do. Behind the 45.5 hours sit three trips that a longer tool life, a back-up tool in the string, or a failure analysis with the tool supplier would attack. They are not a crew problem, and no amount of connection-time discipline touches them.
Equipment NPT and maintenance
Rig repair and waiting on equipment together are 34.5 hours, 36 % of the NPT. The detail is more useful than the code:
| Equipment | Events | Hours |
|---|---|---|
| Top drive | 3 | 13.5 |
| Drawworks brake | 1 | 5.5 |
| Mud pumps | 2 | 4.5 |
| Generator | 1 | 3.0 |
| Waiting on a replacement motor, shaker screens and pump parts | 2 | 8.0 |
The top drive is the largest single item, and it failed on all three rigs. That is the pattern a preventive maintenance system is meant to catch: the same component, the same failure mode, across a fleet. Three practices turn this table into fewer hours next campaign. Code equipment NPT by component as well as by class, so the top drive stands out from "rig repair". Move inspections and service into planned flat time, such as casing runs, cement waits and BOP tests, when the equipment is idle anyway. And keep the parts whose wait cost hours on location: 8 hours here were spent waiting for a motor, shaker screens and pump parts.
The rig's sensors help too. A worn mud pump delivers less flow per stroke well before it fails; a top drive running hot shows in its own channels. Those trends belong with the maintenance plan as well as on the driller's screen.
Connections against the best composite
- Weight to slips: last drilling to slips set
- In slips
- Slips to weight: slips out to drilling again
Every connection measured by DrillIQ's rig-state detector from surface data, 99 per rig. Best composite: the best well average of each part, from different wells (targets_from_offset_wells, best well), 3.91 min; the best single well's whole connection is 4.57 min.
Show the data
| Weight to slips | In slips | Slips to weight | |
|---|---|---|---|
| Rig 1 | 1.08 | 2.47 | 1.17 |
| Rig 2 | 7.08 | 1.76 | 2.67 |
| Rig 3 | 3.00 | 3.57 | 1.17 |
| Best composite | 1.08 | 1.66 | 1.17 |
Rig 1 makes the cleanest connection, 4.72 minutes bottom to bottom, but Rig 2 has the fastest hands in the slips, 1.76 minutes. The best composite takes Rig 1's weight to slips and slips to weight and Rig 2's best well in slips: 1.08 + 1.66 + 1.17 = 3.91 minutes. No well achieved that, yet every part of it was done on this campaign. The best single well's whole connection was 4.57 minutes.
| Rig 1 | Rig 2 | Rig 3 | Target | |
|---|---|---|---|---|
| Connection, min | 4.72 | 11.51 | 7.74 | 3.91 |
| Trip stand moving, min | 1.18 | 1.01 | 1.33 | 0.93 |
| Trip stand in slips, min | 0.75 | 0.53 | 1.17 | 0.50 |
| NPT, h | 29.0 | 31.0 | 36.5 | – |
| ILT, h | 5.2 | 13.6 | 14.6 | – |
A hand check of Rig 2. Against the composite each connection loses (7.08 − 1.08) + (1.76 − 1.66) + (2.67 − 1.17) = 7.60 minutes. Over 99 connections that is 752 minutes, 12.5 hours. The engine sums each connection on its own, so a fast connection cannot offset a slow one, and reports 12.8 hours. Of that, the reaming is 6 minutes × 99 = 9.9 hours.
That is where judgement comes in. Reaming every stand in a lateral may be the price of a clean hole, and a crew that stops doing it to save ten hours can lose a day to a pack-off. Before cutting a practice, check what it buys: if the friction per stand and the cavings stay flat without it on the next well, it was habit. The signs to watch are in our article on the early signs of stuck pipe. The 1.5-minute wait for each survey is worth a conversation with the directional team: many tools can take the survey during the connection and send it while drilling resumes.
Invisible lost time, rig by rig
- NPT, coded on the daily reports
- Invisible lost time: connections and trip stands over the best-well target
Invisible lost time: DrillIQ's minutes over target summed over every connection and trip stand (benchmarks.improvement_deterioration), with the targets from the best offset well by hole size; outliers excluded, the three parts of a connection counted once.
Show the data
| NPT | Invisible lost time | |
|---|---|---|
| Rig 1 | 29.0 | 5.2 |
| Rig 2 | 31.0 | 13.6 |
| Rig 3 | 36.5 | 14.6 |
Across the campaign the engine found 33.3 hours of ILT, against 96.5 hours of NPT: one hour for every three. It is smaller, and it is the cheaper hour to recover, because it needs no new tool and no new rig. It needs each crew to see its own stands against the best ones.
The split differs by rig. Rig 2's 13.6 hours are almost all connections, and most of that is the reaming practice. Rig 3's 14.6 hours are roughly half connections and half trips: its stands wait 1.17 minutes in the slips against a target of 0.50, and over 456 trip stands (0.67 × 456) that is 5.1 hours, which the engine confirms. Rig 1 loses 5.2 hours, mostly on trips, where Rig 2 is quicker.
Two cautions. A target taken from the best offset well is a demonstrated number, and the technical limit lies beyond it; once crews reach it, the stretch target, the best tenth of the best well's stands, takes over. And ILT should be read per hole size. A 6⅛″ lateral connection and a 12¼″ one are different operations, which is why DrillIQ matches targets to offset wells by hole size.
Making it visible on the rig
Dashboards cut NPT and ILT only when they reach the people who can act, in the shift they can act in. Three things make the difference. The rig-state track must be trusted: corrected where the detector is wrong, with NPT entered against it by code. Every stand of the day must be shown against its target, coloured by crew, with the minutes lost added up. And the NPT Pareto must be reviewed by cause across rigs every week while the wells are drilling.
A checklist for the rig team
- Use one NPT code list across every rig and contractor, and add the component for equipment NPT.
- Check that the rig-state track matches what happened before reading any KPI from it.
- Split each connection into weight to slips, in slips and slips to weight; the fix differs for each.
- Set targets from offset wells in the same hole size, and move to stretch targets once they are met.
- Before dropping a practice such as reaming, check what it bought in hole condition.
- Show each crew its own stands against the target at the end of every shift.
- Schedule equipment service into planned flat time, and keep the parts that caused waiting on location.
- Review NPT by cause across rigs weekly; three similar failures are one problem.
How DrillIQ handles it
DrillIQ's KPI & Rig State module classifies every sample of the rig's live or stored data into 17 rig states with a rule-based detector, splits connections and trip stands into their parts, and computes the connection, tripping and flat-time KPIs. Targets come from offset wells matched by hole size, from the best well, its best tenth, the top quartile or the median. The daily, shift and end-of-well reports show every stand against its target, by crew, with the minutes lost summed, and the fleet view ranks rigs by footage per day, NPT and failures.
NPT and corrections are entered against the track by hand or from the daily drilling reports, which the module reads and reconciles; an accepted correction carries into every report. The campaign above was worked out with the same engine.
Book a demo See all capabilities
References
- Bond, D.F., Scott, P.W., Page, P.E. and Windham, T.M. Step Change Improvement and High Rate Learning are Delivered by Targeting Technical Limits on Sub-Sea Wells. IADC/SPE 35077, 1996.
- Maidla, E. and Maidla, W. Rigorous Drilling Nonproductive-Time Determination and Elimination of Invisible Lost Time: Theory and Case Histories. SPE 138804, 2010.
- ISO 14224. Petroleum, petrochemical and natural gas industries: Collection and exchange of reliability and maintenance data for equipment. International Organization for Standardization, 2016.
About this example. The well in this article is a representative scenario set in typical North Oman conditions and worked through with DrillIQ's engines. It is not a client's well, and its figures are not field results.