deepwater horizon investigation - a-o-p · well control response actions failed to regain control...

Deepwater Horizon Accident Investigation 1

Deepwater Horizon Investigation


Investigation Context

Terms of Reference

Investigation Team

Data and Analysis

Investigation Limitations


Well integrity was not established or failed−

Annulus cement barrier did not isolate hydrocarbons

−

Shoe track barriers did not isolate hydrocarbons

Hydrocarbons entered the well undetected and well control was lost−

Negative pressure test was accepted although well integrity had not been established

−

Influx was not recognized until hydrocarbons were in riser

−

Well control response actions failed to regain control of well

Hydrocarbons ignited on the Deepwater Horizon−

Diversion to mud gas separator resulted in gas venting onto rig

−

Fire and gas system did not prevent hydrocarbon ignition

Blowout preventer did not seal the well−

Blowout preventer (BOP) emergency mode did not seal well

1

2

3

4

4

5

Sea Floor

Casing

Riser

BOP

Reservoir1

3

Eight Barriers Were Breached

2

5

6

7

8

6 7

8


Introduction

Technical Presentation


Well Integrity Was Not Established or Failed


Sea Floor

ChokeBoost

BOP

Kill

17:30 –

19:30

Circulated prior

to cement job36”

28”

22”

18”

16”

13-5/8”

11-7/8”

9-7/8”14.17 ppg

SOBM

Bottoms–up Marker

Cement

Mud

Spacer

After drilling to total depth, casing is run to bottom in preparation for the cement job. A double valve float collar is used to prevent backflow or ingress of fluids through the shoe track until the cement hardens and

creates a permanent barrier.

April 18th 00:30 – April 19th 19:30

Long string design robust, consistent with similar wells in the area

9 attempts made to establish circulation to convert float valves

Circulate ~6 times open hole volume, limited circulation due to concerns over creating losses and hole washout

No evidence that hydrocarbons entered the wellbore prior to the cementing operation

Primary reservoir sands

(12.6 ppg)

Production Casing Installation


April 19th 19:30 – April 20th 07:00Nitrogen cement slurry chosen

–

To achieve light weight slurry due to limited pore pressure / fracture gradient window

Possible risk–

Stability of foam–

Relatively small volume–

Susceptible to contamination

Mitigation of risk by–

Thorough testing of slurry design–

Precise placement

Centralization –

6 inline centralizers spaced across the reservoir sands

–

Additional centralizers not run because incorrectly thought to be wrong type

–

Risk of channeling above reservoir sands known and accepted

Cement is pumped down casing through the float collar and up the annulus to isolate the primary reservoir sands.

Sea Floor

ChokeBoost

BOP

Base Oil

Spacer

Cap Cement

Foamed Cement

Tail Cement

Mud

Nitrogen Breakout

36”

28”

22”

18”

16”

13-5/8”

11-7/8”

Shoe –

17,168’

Bottoms–up Marker

Kill

9-7/8”

Shoe –

18,304’

Float Collar –

18,115’ 6 centralizers Primary reservoir sands

(12.6 ppg)

Primary Reservoir

Sands (12.6 ppg)

Centralizer

00:35 –

02:50

Drill-Quip seal assembly installed

and successfully tested.

No lock down sleeve installed.

02:50 –

07:00

Pull out hole with drill pipe.

Set and test casing hanger seal assembly

Pull out of hole with running tool and drill

pipe

Top of

cement 17,260’

Cement Job


Sea Floor

Casing

Riser

BOP

Reservoir

12.6 ppg

Bottom Wiper Plug

Mud

14.17 ppg

Cap

16.74 ppg

Foam

Cement

14.5 ppg

Shoe

Track

Spacer

14.3 ppg

6.7 ppg

Spacer

14.3 ppg

Top Wiper Plug

Bottom Wiper PlugBottom Wiper Plug

Mud

14.17 ppg

Spacer

14.3 ppg

Top Wiper PlugTop Wiper Plug

Bottom Wiper Plug

Spacer14.3 ppg

16.74 ppg

CapCement

FoamCement14.5 ppg

Cement

Channeling

ShoeTrack

Nitrogen Breakout

Cement is pumped down the landing string and casing into the annulus to isolate hydrocarbon

bearing sands.

Foam slurry recommended was a complex design

Risk of contamination using small volume of cement

No fluid loss additives

Incomplete pre-job cement lab testing

Foam slurry was likely unstable and resulted in nitrogen breakout

Base Oil16.7 ppg

Cement Slurry Placement

Float Collar

Tail Cement16.74 ppg

Reamer Shoe Tail Cement

Key Finding #1The annulus cement barrier did not isolate the reservoir hydrocarbons

Centralizers


Cement Slurry Design Issues

50% quality foam at surface conditions was not stable

18.5% quality foam (downhole quality) was not stable

Yield point of the Halliburton slurry was too low for the foam cement (2 lb / 100 ft2

yield point at 135 deg F)

Fluid loss for the base slurry was excessive compared to industry recommendations (302 cc versus 50 cc per 30 min)

Note: QUALITY = Nitrogen Volume / (Nitrogen + Base Slurry Volume)

Unstable Foam Sample

Original Height

Final Height

Cement

An independent lab completed over 500 tests on a representative cement slurry and reported the following:


16ppg Spacer

14.17ppg SOBM (Mud)

8.6ppg Seawater

Influx

Flow Through Shoe Track - Supporting Evidence

CasingShoe

Failure

SealAssembly

FailureYY

YY

YY

YY

YY

YY

YY

YY

NN1400 psi recorded on drill pipe during negative test at 18:30

NNStatic Kill

NNTiming for Gas Arrival to Surface

NNPressure Response from 21:31 to 21:34

NNPressure Increase from 21:08 to 21:14

NNAbility to flow from 20:58

NNRealistic Net Pay Assumption

YYMechanical Barrier Failure Mode Identified

Key Observations for FlowThrough Shoe vs. Seal

Assembly

CasingShoe

Failure

SealAssembly

Failure

YY

YY

YY

YY

YY

YY

YY

YY

NN1400 psi recorded on drill pipe during negative test at 18:30

NNStatic Kill

NNTiming for Gas Arrival to Surface

NNPressure Response from 21:31 to 21:34

NNPressure Increase from 21:08 to 21:14

NNAbility to flow from 20:58

NNRealistic Net Pay Assumption

YYMechanical Barrier Failure Mode Identified

Key Observations for FlowThrough Shoe vs. Seal

Assembly

CasingShoe

Failure

SealAssembly

Failure


Sea Floor

Casing

Riser

BOP

Reservoir

Key Finding #2The shoe track mechanical barriers did not isolate the hydrocarbons

Tail cement is displaced down the casing into the shoe track. The tail cement is

designed to prevent flow from the annulus into the casing. The float collar valves, which provide a second barrier, must close and seal to prevent flow up

the casing.

Shoe track had two types of mechanical barriers: cement in the shoe track and the double check valves in the float collar

Shoe track cement failed to act as a barrier due to contamination of the base slurry by break out of nitrogen from the foam slurry

Hydrocarbon influx was able to bypass the float collar check valves due to either:

Valves failed to convert orValves failed to seal

Flow through shoe confirmed by fluid modeling and Macondo static kill data

Float Collar

Check Valves

Flow Ports

Shoe track cement

Centralizers

Hydrocarbon Flow Path


Hydrocarbons Entered the Well Undetected and Well Control Was Lost


Sea Floor

A positive pressure test verifies the integrity of the casing and seal

assembly.

April 20th 07:00 – 12:00Casing was pressure tested to:

250 psi (low)

2700 psi (high)

Test successful

Proved integrity of blind shear rams, seal assembly, casing and wiper plug

Test does not test the shoe track due to presence of wiper plug

Kill

Cement

Mud

Spacer

2700 psi


Casing (Positive) Pressure Test


ChokeBoost

BOP

Kill

Cement

Mud

Spacer

Seawater

Influx

Sea Floor15:56 –

16:53

424 bbls of 16 ppg

spacer followed by 30 bbls

of freshwater and 352 bbls

of seawater pumped into well

16:54 -

Close Annular

15:04 –

15:56

Seawater pumped into

Boost, Choke, and Kill lines

16:54 –

16:59

50 bbls bled off

drill pipe due to

leaking annular


(12.6 ppg)

April 20th 15:04 – 19:55Negative test simulates underbalanced condition

Spacer used between mud and seawater

Leaking annular at start of test moved spacer across kill line inlet

Negative test started on drill pipe but changed to kill line

Bleed volumes higher than calculated

Drill pipe built pressure to 1400 psi with no flow on the kill line

The negative-pressure test checks the integrity of the shoe track, casing and

wellhead seal assembly. This simulates conditions during temporary

abandonment when a portion of the well is displaced to seawater.

Negative Pressure Test


ChokeBoost

BOP

Kill

Cement

Mud

Spacer

Seawater

Influx

Sea Floor

Cement Tank

Total

Volume

15 bbls


(12.6 ppg)

17:27Bled 15 bbls of

seawater from drill pipe

Decision made to

change test to kill line










Negative Pressure Test17:52 –

18:00

Open kill

line to conduct

negative test

Bled 3 -

15 bblsinto kill line

Flow did not

stop and

“spurted”

Kill line closed

16:59 –

17:08

Annular seals with

increased hydraulic

closing pressure

Fill riser with 50 bbls of mud

17:08 –

17:27

Monitored that the

annular sealed


ChokeBoost

BOP

Cement

Mud

Spacer

Seawater

Influx

Sea Floor

1400psi

Cement Tank

Total

Volume

15 bbls18 bbls

0

psi

Additional

3 bbl influx

18:00 –

18:35

Drill pipe pressure gradually

increased to 1400 psi


(12.6 ppg)










Negative Pressure Test

18:42 –

19:55

Monitored kill line for 30 min

1400 psi on drill pipe described

as a “bladder effect”

19:55

Negative pressure test

was concluded and

considered a good test

18:42Pumped into kill

line to confirm full

Kill line opened for

monitoring negative test


Sea Floor

Shoe – 18,304’

FC – 18,115’

TOC – 17,260’

Shoe – 17,168’

KillChokeBoost

SOBM

Spacer

Seawater

Influx

SOBM

Spacer

Seawater

Influx

BOP

1400

PSI0

PSI

Key Finding #3The negative pressure test was accepted although well integrity had not been established

Spacer

Sea Floor

Casing

Riser

BOP

Reservoir

Bleed volumes not recognized as a problem

Anomalous pressure on drill pipe with no flow from kill line

Test incorrectly accepted as successful

Negative testing not standardized


Well Monitoring – Driller’s Console and Mudlogging unit

Well monitoring is performed to understand if the well has losses or gains

Driller is responsible for monitoring and shutting in the well

The mudlogger provides monitoring support to the driller

Displays and trending capability available in both Driller’s and Mudlogger’s cabins

Flow, pressure and pit sensors can indicate flow

Simultaneous activities were taking place on April 20th to prepare for rig move

Standards for monitoring do not specifically address end-of-well activities


Choke

Boost

Kill

Cement

Mud

Spacer

Seawater

Mud + Seawater Mix

Influx

Sea Floor

20:00 –

21:08

Resumed pumping

Displaced riser with seawater

until spacer is at surface

21:08

Spacer arrived at surface

Shut pumps down

for sheen test

20:58 -

21:08

39 bbl gain

20:02

Annular opened

after negative

test

20:52

Well becomes underbalanced


(12.6 ppg)

BOP

Mud in the riser is displaced with seawater in preparation for temporary

abandonment.

April 20th 19:55 – 21:14

20:02 Resume displacement of mud with seawater

20:52 Well becomes underbalanced and starts to flow

After 20:58 gain being taken and pressure begins increasing

–

Flow from well masked by emptying of trip tank

21:08 Pumping stops for sheen test

–

Pressure increases with pump off

21:14 Sheen test complete, displacement resumes

Undetected Flowing Conditions


Flow Indications

0

200

400

600

800

1000

1200

1400

1600

1800

2000

20:4

5

20:5

0

20:5

5

21:0

0

21:0

5

21:1

0

21:1

5

21:2

0

21:2

5

21:3

0

21:3

5

Flow

Rat

e (g

pm)

0

500

1000

1500

2000

2500

3000

Pum

p Pr

essu

re (p

si)

Flow Out (calibrated)Flow In (rig pumps)DP Press (rig pumps)

Key Finding #4The influx was not recognized until hydrocarbons were in the riser

Flow indications:

#1: Drill pipe pressure increased by 100 psi, (expected decreased); ~39 bbl gain from 20:58 to 21:08

21:08

1,017 psi

SOBM (mud)

Seawater

Influx

SOBM + seawater mix

39 bbl

300 bbl

Cumulative Gain

Indication #1

0 bbl

20:52-Flow starts

Decreasing trend should have continued

Based on Real-time Data


Flow Indications

0

200

400

600

800

1000

1200

1400

1600

1800

2000

20:4

5

20:5

0

20:5

5

21:0

0

21:0

5

21:1

0

21:1

5

21:2

0

21:2

5

21:3

0

21:3

5

Flow

Rat

e (g

pm)

0

500

1000

1500

2000

2500

3000

Pum

p Pr

essu

re (p

si)



Flow indications:


#2: Drill pipe pressure increased by 246 psi with pumps off

–

Flow out does not immediately drop after shutting down pump

21:08

1,017 psi

SOBM (mud)

Seawater

Influx

SOBM + seawater mix

39 bbl

300 bbl

Cumulative Gain

Indication #1

Indication #2

0 bbl

20:52-Flow starts

Decreasing trend should have continued

Overboard line openedFlow out available only

to driller after 21:10


Normal Flow Back

0

200

400

600

800

1000

1200

16:5

0

16:5

5

17:0

0

17:0

5

Flow

Rat

e (g

pm) Flow Out

Flow In


Flow Indications

0

200

400

600

800

1000

1200

1400

1600

1800

2000

20:4

5

20:5

0

20:5

5

21:0

0

21:0

5

21:1

0

21:1

5

21:2

0

21:2

5

21:3

0

21:3

5

Flow

Rat

e (g

pm)

0

500

1000

1500

2000

2500

3000

Pum

p Pr

essu

re (p

si)



Flow indications:


#2: Drill pipe pressure increased by 246 psi with pumps off

–

Flow out does not immediately drop after shutting down pump

#3: Drill pipe pressure increased by 556 psi with pumps off; ~300 bbl gain

No well control actions taken

21:08

1,017 psi

21:31

1,200 psi

SOBM (mud)

Seawater

Influx

SOBM + seawater mix

39 bbl

300 bbl

Cumulative Gain

Indication #1

Indication #2

0 bbl

Indication #3

20:52-Flow starts



0

500

1000

1500

2000

2500

3000

21:3

0

21:3

2

21:3

4

21:3

6

21:3

8

21:4

0

21:4

2

21:4

4

21:4

6

21:4

8

21:5

0

Drill Pipe Presssure (psi)

Pumps shut down

Discussion about “Differential Pressure”

Attempt to bleed pressure

Close Drill Pipe

Mud overflowing onto rig floor

BOP Sealing

Pressure increase due to annular activation

BOP

Key Finding #5Well control response actions failed to regain control of the well

Exp

losi

on

at

21:4

9

-

Mud and water raining onto deck

-

TP calls WSL, getting mud back, diverted to MGS, closed or was closing annular

-

AD calls Senior TP, Well blowing out, TP is shutting it in now

Mud shoots up derrick-Diverter closed-BOP activated

First indication of well control response:49 minutes and 1000 bbls after initial influx

Influx enters riser

Annular leaking



Hydrocarbons Ignited on the Deepwater Horizon


When responding to a well control event the riser diverter is closed and fluids

sent to either the mud gas separator or to the overboard diverter lines.

KillBoost

Choke

BOP

14”

Diverter LinePort

Overboard 14”

Diverter Line

Mud

System

Starboard Overboard

6”

Vacuum

Breaker

Rated to 60 psi

working pressure

IBOP

Slip Joint

Rotary

Hose

Diverter

12”

Vent

MGS

Bursting Disk

Starboard Overboard

Rated to 100 or 500 psiOverboard

Caisson

Diversion to the MGS

Rig crew has the option to divert flow to port/starboard overboard lines or the MGS

Diverting to port or starboard will result in fluids venting overboard

Liquid outlet from MGS goes to the Mud System under the main deck

Seawater

Seawater/Mud Mix

Influx

Diverting to the Mud Gas Separator at about 21:42


Hydrocarbon flow from surface equipment

Instantaneous gas rates reached 165 mmscfd

Pressures exceeded operating ratings (above 100 psi)

Gas would probably have vented from:

Slip joint packer into the moon pool

12” MGS “gooseneck” vent

6” MGS vacuum breaker vent

6” overboard line through burst disk

10” mud line under the main deck

When responding to a well control event the riser diverter is closed and fluids

sent to either the mud gas separator or to the overboard diverter lines.

KillBoost

Choke

BOP

14”

Diverter LinePort

Overboard 14”

Diverter Line

Mud

System

Starboard Overboard

6”

Vacuum

Breaker

IBOP

Rotary

Hose

Diverter

12”

Vent

MGS

Bursting Disk

Starboard Overboard

Overboard

Caisson

Slip Joint

Seawater

Seawater/Mud Mix

Influx

Gas flow to Surface at high rate: 21:46 to 22:00


Animation of Gas Dispersion

Upper Explosive Limit

Lower Explosive Limit

Cut Section Through Derrick Towards Aft3D View

Gas Dispersion across the Deepwater Horizon 21:46 to 21:50 hrs


Aft

Fwd

3D view

Secondary Protective Systems

Gas cloud reached the supply air intakes for engine rooms 3, 4, 5 & 6

The Fire and Gas system did not automatically trigger a shutdown of the HVAC system for the engine rooms

Limited areas of the rig are designated as electrically classified zones

Secondary protective systems are designed to reduce the potential

consequence of an event once the primary protective systems have failed.

Secondary protective systems did not prevent ignition


Hydrocarbons were routed to the mud gas separator instead of diverting overboard

Resulted in rapid gas dispersion across the rig through the MGS vents and mud system

Key Finding #6Diversion to the mud gas separator resulted in gas venting onto the rig

When responding to a well control event the riser diverter is closed and fluids are sent to either the mud gas separator or

to the overboard diverter lines.

KillBoost

Choke

BOP

14”

Diverter LinePort

Overboard 14”

Diverter Line

Mud

System

Starboard Overboard

6”

Vacuum

Breaker

IBOP

Rotary

Hose

Diverter

12”

Vent

MGS

Bursting Disk

Starboard Overboard

Overboard

Caisson

Slip Joint

Seawater

Seawater/Mud Mix

Influx

BOP Sealed at

21:47


Key Finding #7The fire and gas system did not prevent hydrocarbon ignition

Gas dispersion beyond electrically classified areas

Gas ingress into engine rooms via main deck air intakes

The on-line engines were one potential source of ignition

Aft

Fwd3D view

Aft

Aft

Gas Dispersion at 4 minutes

Fwd3D view

Section through derrick

(Upper Explosive Limit)

(Lower Explosive Limit)

Secondary protective systems are designed to reduce the potential

consequence of an event once the primary protective systems have

failed.


Emergency Well Control System Did Not Seal the Well


Blowout Preventer (BOP)

Surface HPU & Accumulators

Lower Stack Accumulators

Mux Cable

Blue Control

Pod

Yellow Control

Pod

Hydraulic Conduit

LMRP Accumulators

BOP Control Panel

Manual Automatic ROV Intervention

EDS

HP BSR Close AMF

HOT StabAMF

Auto-shear

Emergency Methods of BOP Operation Available on DW Horizon

Sea Bed

Wellhead Connector

Wellhead

Upper Annular

Lower Annular

Stripping Element

Casing Shear Ram

(Non Sealing)

Upper VBR

Middle VBR

Lower (Test) VBR

Blind Shear Ram

Flex Joint LMRP

BOP Stack

Deepwater Horizon Accident Investigation 33Sea Bed

Wellhead Connector

Wellhead

Upper Annular

Lower Annular

Stripping Element

Casing Shear Ram

(Non Sealing)

Upper VBR

Middle VBR

Lower (Test) VBR

Blind Shear Ram

BOP is designed to seal the wellbore and shear casing or drill pipe if necessary.

April 20th

21:41 annular BOP closed but appears not to have sealed the annulus

21:47 a VBR likely closed and sealed the annulus

Activation of

Lower Annular BOP

Activation of VBR

April 20th

21:38 –

Hydrocarbons

enter the riser

BOP Response (Before the Explosions)


Wellhead Connector

Wellhead

Upper Annular

Lower Annular

Stripping Element

Casing Shear Ram

(Non Sealing)

Upper VBR

Middle VBR

Lower (Test) VBR

Blind Shear Ram

MUX cables provide electronic communication and electrical power to

the BOP control pods.

Annular BOP

gradually opens

BOP Response (Impact of Explosions)

April 20th

Damage to MUX cables and hydraulic line

–

Opening of annular BOP

Rig drifted off location

–

Upward movement of the drill pipe in the BOP


April 20th

EDS attempts failed to activate BSR

AMF sequence likely failed to activate BSR

April 21st – 22nd

ROV hot stab attempts to close BOP were ineffective

ROV simulated AMF function likely failed to activate BSR

ROV activated auto-shear appears to have activated but did not seal the well

April 25th – May 5th

Further ROV attempts using seabed deployed accumulators were unsuccessful

Wellhead Connector

Wellhead

Upper Annular

Lower Annular

Stripping Element

Casing Shear Ram

(Non Sealing)

Upper VBR

Middle VBR

Lower (Test) VBR

Blind Shear Ram

BSR activated byAuto-shear

BOP Response (After the Explosions)

There are several emergency methods of activating the BSR to seal the well.


AMF

Key Finding #8The BOP emergency mode did not seal the well

EDS function was inoperable due to damage to MUX cables

AMF could not activate the BSR due to defects in both control pods

Auto-shear appears to have activated the BSR but did not seal the well

Potential weaknesses found in the BOP testing regime and maintenance management systems

The AMF provides an automatic means of closing the BSR without

crew intervention.

Manual Automatic ROV Intervention

EDS

HP BSR Close

Emergency Methods of BOP Operation Available on DW Horizon

HOT StabAMF

Auto-shear

Damaged MUX Cable

Blue Control

Pod

Yellow Control

Pod

Damaged Hydraulic Conduit

Explosions & Fire: Loss of communicationLoss of electrical power

Loss of hydraulics


Summary of Findings and Recommendations


Recommendations25 Recommendations Specific to the 8 Key Findings

BP Drilling Operating Practice and Management SystemsEngineering Technical Practices and Procedures

Further Enhance Deepwater Capability and Proficiency

Strengthen Rig Audit Action Closeout and Verification

Introduce Integrity Performance Management for Drilling and Wells Activities

Contractor and Service Provider Oversight and AssuranceCementing Services

Drilling Contractor Well Control Practices and Proficiency

Oversight of Rig Safety Critical Equipment

BOP Configuration and Capability

BOP Minimum Criteria for Testing, Maintenance, System Modifications and Performance Reliability

BP has accepted all the recommendations and is reviewing how best to implement across its world wide operations


Well integrity was not established or failed−

Annulus cement barrier did not isolate hydrocarbons

−

Shoe track barriers did not isolate hydrocarbons

Hydrocarbons entered the well undetected and well control was lost−

Negative pressure test was accepted although well integrity had not been established

−

Influx was not recognized until hydrocarbons were in riser

−

Well control response actions failed to regain control of well

Hydrocarbons ignited on the Deepwater Horizon−

Diversion to mud gas separator resulted in gas venting onto rig

−

Fire and gas system did not prevent hydrocarbon ignition

Blowout preventer did not seal the well−

Blowout preventer (BOP) emergency modes did not seal well

1

2

3

4

4

5

Sea Floor

Casing

Riser

BOP

Reservoir1

3

Summary of Key Findings

2

5

6

7

8

6 7

8

deepwater horizon investigation - a-o-p · well control response actions failed to regain control...

Documents