HILCORP ALASKA, LLC

Incident Narrative

AT APPROXIMATELY 16:30 HOURS ON APRIL 1, 2021 DURING NORMAL HELICOPTER OPERATIONS BUBBLES WERE OBSERVED AT THE SURFACE OF THE WATER. NRC WAS NOTIFIED LESS THAN ONE HOUR LATER. THE LOCATION OF THE BUBBLES WAS MAPPED AND CONFIRMED TO BE WITHIN CLOSE PROXIMITY TO MGS A PIPELINE. PRODUCTION ON THE PLATFORM WAS SHUT-IN AND PIPELINE PRESSURE WAS REDUCED TO LIMIT THE LEAK TO THE FULLEST EXTENT POSSIBLE WHILE MAINTAINING POSITIVE PRESSURE INSIDE THE PIPELINE. LOCAL DIVE RESOURCES WERE CONTACTED AND CONTRACTED TO BEGIN GATHERING AND PLANNING FOR PIPELINE REPAIRS. ON APRIL 2, 2021 THE PIPELINE PRESSURE WAS LOWERED TO 64 PSIG. ARRANGEMENTS WERE MADE TO MAINTAIN ELECTRICAL GENERATION AND HEAT UTILIZING DIESEL GENERATORS. SUBSEA SONAR RESOURCES WERE ENGAGED AND BEGAN PREPARING FOR A MULTI-BEAM SONAR SURVEY OF THE AREA TO VERIFY THE LEAK AND DETERMINE THE LEAK LOCATION. MONITORING OF ICE FLOWS AND WEATHER CONDITIONS BEGAN FOR THE AREA, SO THE SUBSEA RESOURCES COULD BE SAFELY MOBILIZED. ON APRIL 3, 2021 OPERATIONS SHUT IN MGS A GAS PIPELINE. HILCORP POSITIVELY IDENTIFIED THE LEAK LOCATION ON APRIL 6, 2021. DIVERS HAVE BEEN MOBILIZED TO THE SITE AND HILCORP IS CURRENTLY WORKING TO DETERMINE THE CAUSE OF THE GAS LEAK IN ACCORDANCE WITH THE AMENDED CORRECTIVE ACTION ORDER ISSUED BY PHMSA ON APRIL 6, 2021. THE DAMAGED PIPELINE SEGMENT WAS REMOVED FOR FURTHER INVESTIGATION AND PIPELINE ENDS STABILIZED. ON AUGUST 27, 2021 ADV INTEGRITY ISSUED A ROOT CAUSE FAILURE ANALYSIS TO HILCORP AS FOLLOWS: 1. THE SUBJECT FAILURE OCCURRED DUE TO A TIME-DEPENDENT CRACKING MECHANISM, LIKELY FATIGUE, WHICH WAS LIKELY EXACERBATED BY OTHER ENVIRONMENTAL FACTORS. THE FEATURES IDENTIFIED WERE NOT CONSISTENT WITH TYPICAL PIPELINE STRESS CORROSION CRACKING. 2. THE LOCAL STRESS STATE CONTRIBUTED TO THE OBSERVED THROUGH-WALL CRACK. THIS LOCAL STRESS STATE INCLUDED: (1) THE HIGH TENSILE LOAD RESULTING FROM THE PIPE'S CONTOUR ACROSS THE ROCKY OUTCROPPING, (2) THE PRESENCE OF THE PLIDCO CLAMP IMMEDIATELY ADJACENT TO THE FAILURE WHICH CONTRIBUTED TO THE RIGIDITY OF THE PIPELINE, AND (3) THE REDUCED WALL THICKNESS AS A RESULT OF EXTERNAL CORROSION AND ABRASION. 3. MECHANICAL PROPERTIES OF THE PIPE AND GIRTH WELD WERE CONSISTENT WITH API 5L, GRADE B MATERIAL. SUBPART K SELECTION DESCRIPTIONS: - PREVIOUS MECHANICAL DAMAGE NOT RELATED TO EXCAVATION: THE PREVIOUS OWNER INSTALLED A PLIDCO PIPE CLAMP IN 2014 JUST UPSTREAM OF THE FAILURE LOCATION. IMPROPER SUPPORT OF THE PIPELINE DURING AND AFTER LIFTING COULD ALSO IMPART RESULTING STRESS AT AND NEARBY THE PLIDCO. - EXTERNAL CORROSION, GALVANIC: REDUCED WALL THICKNESS DUE TO EXTERNAL CORROSION AND ABRASION WHERE FOUND AT THE SITE OF THE LEAK. - INSTALLATION RELATED: INCREASED TENSILE LOAD RESULTING FROM THE PIPE'S CONTOUR ACROSS THE ROCKY OUTCROPPING, AND THE INCREASED RIGIDITY FROM THE NEAR BY PLIDCO CLAMP INCREASING THE STRESS CONCENTRATION FACTOR IN THAT LOCATION. THIS PIPELINE WILL NOT BE BROUGHT BACK INTO SERVICE AS HILCORP HAS DECIDED TO PURSUE THE ABANDONMENT OF THIS LINE. NOTE: THE ORIGINAL REPORT SUBMITTED ON 4/29/21 ERRONEOUSLY REPORTED 1,052 MCF UNINTENTIONAL AND 1,567 MCF INTENTIONAL RELEASE VOLUMES DUE TO CONFUSION OVER THE DEFINITION OF THESE TWO METRICS. 2,619 MCF OF GAS WAS UNINTENTIONALLY RELEASED FROM THE TIME THE LEAK BEGAN UNTIL THE SHUT-IN PIPELINE PRESSURE EQUALIZED WITH THE ATMOSPHERIC PRESSURE AT THE BOTTOM OF THE INLET. CONTROLLED AND INTENTIONAL DISPLACEMENT OF GAS USING NITROGEN WAS USED TO DETERMINE THE EXACT LOCATION OF THE LEAK. MULTI-BEAM SONAR FIRST CONFIRMED THE LOCATION. THEN DIVERS WERE MOBILIZED TO INSPECT THE LOCATION. THE PIPELINE WAS AT 35 PSIG (STATIC HEAD PRESSURE) WHEN LEAK LOCATION OPERATIONS WERE ENACTED. AT 35 PSIG THE TOTAL PIPELINE VOLUME IS ESTIMATED AT 34 MCF OF NATURAL GAS. THEREFORE HILCORP IS ASSUMING A MAXIMUM INTENTIONAL GAS RELEASE OF 34 MCF.