The end-of-well report is written once, when the rig has gone and most people have moved on. Its real reader is the team planning the next well, and for them a list of events is not enough: they need to know which days were lost, why, and which of them the next programme can take back. This article works through the report for a horizontal well, from the depth-against-days curve to invisible lost time, and turns it into the next well's programme and targets.
What the report is for
Most end-of-well reports answer "what happened". The useful ones answer three more questions. Where did the well lose time against its programme? How much of that loss was visible, coded as non-productive time (NPT) on the daily reports, and how much was hidden inside work that looked productive? And what should change on the next well, written as a number someone can be held to?
The answers come from three records the well already produced. The drilling programme gives the plan, section by section. The daily drilling reports give the depth at the end of each day and the hours coded as NPT. The rig's sensor data, cut into rig states and stands, gives every connection, tripping stand and casing joint as a measured duration. Put together, they explain every day of the well.
A report built for the next well usually runs in this order: a one-page summary with the days, cost and NPT against the programme and AFE; the depth-against-days curve; one page per section with its plan, actual, NPT and lost time; the KPI table against targets; cost by section and cost per metre; a short geology and directional summary (formation tops against prognosis, the final survey, targets hit); and the lessons, each with an owner. The first page should let a manager who reads nothing else see whether the well delivered and why. Everything after it is evidence for the lessons, so each page should end with what it means for the next well.
The method: three layers of lost time
Split each section's overrun against the programme into three parts:
ΔT = Tactual − Tplan = NPT + ILT + R
ILT = Σ max(0, ti − ttarget)
- NPT
- hours coded as non-productive on the daily reports: losses, equipment failure, waiting
- ILT
- invisible lost time: for each connection, tripping stand or casing joint i, the minutes it took over the target for that operation in that hole size
- ttarget
- the best average any offset well achieved for that operation and hole size
- R
- the rest: drilling slower or faster than planned, flat time over or under its estimate
Targets are matched by hole size, not by section name, because the names differ from well to well and the bit size does not. A target taken from the best offset's average is a level the same rig has already reached. A stretch target, the best tenth of the best offset, is one it has reached on a good day.
The residual matters as much as the other two. A section can carry hours of NPT and still finish early because drilling was faster than planned. That speed is a lesson too.
The example well
The example is a horizontal oil producer in a Lower Cretaceous carbonate reservoir, set in typical onshore Abu Dhabi conditions. It is the third well drilled from its pad by the same rig, so two earlier wells there serve as offsets.
| Section | Hole and casing | Programme | Actual | NPT |
|---|---|---|---|---|
| Rig move | between pads, rig up | 3.0 d | 3.0 d | |
| Surface | 17½″ to 700 m, 13⅜″ casing | 4.5 d | 5.0 d | 20.4 h, total losses at 420 m |
| Intermediate | 12¼″ to 2,250 m, wireline log, 9⅝″ casing | 10.5 d | 10.0 d | |
| Build | 8½″ to land at 2,800 m, 7″ liner | 7.0 d | 8.0 d | 16.8 h, rotary steerable failure at 2,560 m |
| Drain | 6″ to 4,000 m | 6.5 d | 7.0 d | |
| Completion | lower and upper completion, rig down | 5.5 d | 6.0 d | |
| Well | 37.0 d | 39 d | 37.2 h |
The section days come from DrillIQ's plan-against-actual engine, reading the programme and the 39 daily reports: each section runs from drilling out to the next section's drill-out, the last one to TD. The 2.0 days the well ran over is the number the report has to explain.
The depth curve
- Programme (published plan)
- Depth at 24:00 on each daily report
DrillIQ's plan-against-actual engine (Daily Ops Reports, shared_plan.plan_vs_actual) on the example's programme and its 39 daily reports. NPT is the hours coded as NPT on the reports.
Show the data
| Day | Programme | Daily reports |
|---|---|---|
| 0.0 | 30 | 30 |
| 3.0 | 30 | 30 |
| 5.5 | 700 | 511 |
| 7.5 | 700 | 700 |
| 14.5 | 2,250 | 2,190 |
| 18.0 | 2,250 | 2,250 |
| 22.5 | 2,800 | 2,685 |
| 25.0 | 2,800 | 2,800 |
| 31.5 | 4,000 | 3,743 |
| 37.0 | 4,000 | 4,000 |
The depth-against-days curve is the report's first page because it shows the shape of the well at a glance. Vertical steps are drilling; flat runs are casing, logging, completion and lost time. The two curves match closely through the 12¼″ hole, which drilled half a day faster than planned. The losses at 420 m push the actual curve right in the surface hole; the rotary steerable failure at 2,560 m pushes it right again in the build. TD came on day 33 against the programme's 31.5.
The curve shows where time went but not why. A flat run on the curve looks the same whether the rig was waiting on cement or tripping slowly. For that the report needs the sensor data.
Visible and invisible lost time
- NPT, coded on the daily reports
- Invisible lost time: connections, trips and casing over the best offset well
Invisible lost time from DrillIQ's KPI engine (benchmarks.improvement_deterioration): each connection, tripping stand and casing joint against the best of the two earlier wells on the pad, by hole size (targets_from_offset_wells).
Show the data
| NPT (coded on the reports) | Invisible lost time (over the best offset) | |
|---|---|---|
| 17½″ surface | 20.4 | 1.9 |
| 12¼″ intermediate | 0.0 | 4.5 |
| 8½″ build | 16.8 | 7.9 |
| 6″ drain | 0.0 | 4.9 |
NPT is the lost time everyone already knows: 20.4 hours curing losses in the surface hole and 16.8 hours for the rotary steerable failure, 37.2 hours in all. The invisible lost time is 19.1 hours: 12.7 in trips, 3.7 in connections and 2.8 running casing. It is the sum, over every one of the well's 138 connections, 810 tripping stands and 301 casing joints, of the minutes each took over the best offset's average for its hole size.
The build section carries the most of both. Its 7.9 hours of ILT come mostly from trips: 360 stands, half of them from the round trip the tool failure forced, with each stand 0.56 minutes slower in slips and 0.51 minutes slower moving than the offsets' best. A failure costs its coded NPT and then again in the extra pipe handling it brings, and only the first shows on the daily report.
A hand check
- The build's trips, in slips. This well averaged 1.74 minutes per stand against the best offset's 1.18, over 360 stands: (1.74 − 1.18) × 360 = 202 minutes, 3.4 hours. The engine, stand by stand, finds 209 minutes over the target and 9 minutes under it: 200 minutes net. The chart counts only the time over, 3.5 hours.
- The well's 2.0 days. NPT is 37.2 / 24 = 1.55 days and ILT 19.1 / 24 = 0.80 days, 2.35 days together. The residual is −0.35 days. Once their lost time is taken off, the surface and 12¼″ sections ran 1.12 days faster than planned and the build 0.03 days faster, while the drain ran 0.30 and the completion 0.50 days slower. 1.55 + 0.80 − 0.35 = 2.0.
- What the lost time cost. At a time-related spread rate of 45,000 USD a day, assumed here for rig, services and rentals, NPT cost 1.55 × 45,000 = 69,750 USD and ILT 0.80 × 45,000 = 35,900 USD.
The ILT is the smaller number, but it is the one most within the rig team's control. NPT events are different every well. Slow pipe handling repeats on every stand of every trip until someone changes it.
From this well to the next programme
- Programme
- Actual
- Actual less its NPT and invisible lost time
Section days from drilling out to the next section's drill-out (the last section to TD), from the plan-against-actual engine; achievable = actual − NPT − invisible lost time.
Show the data
| Programme | Actual | Actual less NPT and invisible lost time | |
|---|---|---|---|
| 17½″ surface | 4.5 | 5.0 | 4.1 |
| 12¼″ intermediate | 10.5 | 10.0 | 9.8 |
| 8½″ build | 7.0 | 8.0 | 7.0 |
| 6″ drain | 6.5 | 7.0 | 6.8 |
Take each section's actual days, remove its NPT and its ILT, and what is left is what this rig and crew have shown they can do. For the surface hole that is 4.1 days, faster than its own 4.5-day plan; for the 12¼″ hole 9.8 days against 10.5; for the build 7.0 days, on plan. The drain is the exception: 6.8 days, slower than the 6.5 planned even without lost time, so its plan stays and its rate of penetration goes on the list of things to investigate.
A next programme built that way, taking the smaller of plan and achievable for each section and keeping the rig move and completion as planned, is 3.0 + 4.1 + 9.8 + 7.0 + 6.5 + 5.5 = 35.9 days. That is 1.1 days under this well's programme and 3.1 days under what it took. The surface losses do not go into the firm plan: they become a contingent step of 0.85 days, with the losses material on location, so the plan stays honest and the risk stays visible. For how a programme is built from offset wells in the first place, and how the AFE follows it, see the AFE article.
The KPI targets go with it. Because this well beat neither offset's average in any operation, the best-well targets for the next well are the same as this well's: weight to weight 9.41, 6.73, 8.50 and 7.02 minutes in the 17½″, 12¼″, 8½″ and 6″ holes, tripping stands 1.18 to 1.35 minutes in slips and 1.98 to 2.22 minutes moving, casing joints 2.60 to 3.09 minutes in slips. Against its own best tenth instead, the well left 26.3 hours on the table, 7.6 of them in the build's trips. That measure needs no offsets at all: the crew has already done it.
Lessons that change the next well
A lesson earns its place when it names the cause, the action, the owner and where it goes in the next programme. Everything else is a history.
| What happened | Action on the next well | Where it goes |
|---|---|---|
| Total losses at 420 m, 20.4 h | LCM in the system before 400 m; plug material and a cement plan on location | surface-hole programme; contingent step 0.85 d |
| Rotary steerable failure, 16.8 h and a round trip | review the tool's run hours and failure report; backup tool on location for the build | BHA plan; service-quality review |
| Build trips 0.5 min per stand slower than the best offset, in slips and moving | trip-speed and slips-time targets per hole size on the rig floor | KPI targets |
| Connections 0.8 to 1.9 min slower than the best offset | a connection routine agreed with both crews, tracked stand by stand | KPI targets; crew briefings |
| Drain 0.3 d slower than planned after lost time | bit and parameter review against offset drains | drain programme; bit plan |
Two habits make the list reliable. Write the report within a week of release, while the people who were there still remember. And close the loop: at the next well's pre-spud meeting, read the lessons aloud and point to where each one sits in the new programme. The NPT side of this analysis, and how to code it consistently so it can be compared well to well, is covered in understanding NPT; the daily reports it rests on, in automating daily drilling reports.
A checklist for the well team
- Compare every section against the programme, in days, from drill-out to drill-out.
- Split each overrun into NPT, invisible lost time and the rest; make the three add up.
- Measure connections, trips and casing runs against offset targets by hole size, not by section name.
- Count a failure's extra pipe handling with the failure, not only its coded NPT.
- Build the next programme from what this well showed is achievable, section by section.
- Carry known hazards as contingent steps, not as padding in the firm plan.
- Write every lesson with its cause, action, owner and place in the next programme.
- Read the lessons at the next pre-spud meeting.
How DrillIQ handles it
DrillIQ builds the end-of-well report from records the well already holds. The KPI & Rig State module's End of Well report draws the depth against days on the published programme's own curve, the time by section, every KPI against its target with the hours lost, the time gained and lost, trends against depth, every stand coloured by crew, and the well against its own best tenth in hours and at the rig rate. Targets come from offset wells by hole size, as best well, stretch, top quartile or median, previewed before they are applied. The report builder turns pages on and off, adds the engineer's remarks and marks the report draft, reviewed or issued.
Daily Ops Reports compares the daily reports with the programme section by section, with the NPT in each; the AFE module gives the cost and cost per metre by section and keeps a variance register with the root cause and lesson for each overrun; the Well Construction Plan carries contingent steps beside the firm plan. Everything above was worked out with the same engines.
Book a demo See all capabilities
References
- Adams, N.J. Drilling Engineering: A Complete Well Planning Approach. PennWell, 1985.
- Mitchell, R.F. and Miska, S.Z. (eds.) Fundamentals of Drilling Engineering. SPE Textbook Series Vol. 12, Society of Petroleum Engineers, 2011.
- ISO 14224:2016. Petroleum, petrochemical and natural gas industries: Collection and exchange of reliability and maintenance data for equipment.
About this example. The well in this article is a representative scenario set in typical onshore Abu Dhabi, UAE conditions and worked through with DrillIQ's engines. It is not a client's well, and its figures are not field results.