Home Search Emissions Pollutants About the Database

Marathon Ashland Petroleum (3165), Garyville

Releases of Highly Reactive Volatile Organic Compounds

LDEQ Accident Number
Accident Date
Point Source(s) Notes Amount of Release

FCC Unit 205
Cause: Due to an unexpected loss of power to the control system of the Unit 25 FCCU. FCCU shut down as designed which resulted in less gas feed to the Unit 205 Coker. This decreased in feed caused the Coker Wet Gas Compressor Suction Drum to briefly exceed the maximum safe operating pressure of the drum which resulted in the Coker Wet Gas relieving tot he ground flare. There was no known offsite impacts resulting from this incident. The emissions from the FCCU shut down are permitted as part of the overall North Flare. Compressor spillback opened rapidly to compensate and a high pressure was reached on the suction drum. Pressure control valve opened to flare once pressure reached 21 psig.

Followup: No

Notes: The FCCU was safely shut down and all other related unit feed rates were adjusted per the FCCU shutdown plan. The Coker Unit Wet Gas Compressor control system compensated for the increased suction pressure by increasing the compressor speed. All aspects of this incident are currently under investigation.
0.6 pounds


Unit 59 North Flare
Cause: The initiating incident was a pump seal fire in the Gasoline Desulfurization Unit (Unit 55). The fire was fueled by a leaking seal on the pump. Extinguishing the fire was delayed by inability to close an EIV on the suction side of the pump. This resulted in emergency shutdown of the unit. Two other events also occurred on this day including an upset in Sulfur Plant Unit 234 and a flame-out of the North Flare. Due to the fire and emergency shutdown of the Gasoline Desulfurization Unit, the Fluid Catalytic Cracking Unit cut feed, sending vent gas to the North Flare. Process vent gas was sent to the North Flare which increased the steam to the flare suddenly, snuffing the flare out.

Followup: Yes

Notes: PDF too large to upload (109 pages) To re-light the North Flare, steam was gradually decreased and natural gas was added to the flare gas to allow the two available igniters to relight the North Flare. Parts to repair the North Flare pilot system were already on order when this incident occurred. The North Flare was taken out of service when the parts were received and repaired on October 31, 2012. Spare pilot and igniter assemblies are now in stock so that repairs can be made in a timely fashion if an incident like this is to occur again. Total amount of pollutants released was 59438.44 lbs, but 90% was claimed to be efficiently burned off, resulting in 5943.59 lbs that were actually released. The reportable quantity for Highly Reactive Volatile Organic Compounds (HRVOCs) (100 pounds) was exceeded during the 24 hour period.
105.2 pounds

outfall 002
Cause: There were multiple units that experienced upsets during the shutdown and startup activities surrounding Hurricane Isaac: In preparation for Hurricane Isaac, the refinery units were at minimum rates anticipating a shutdown condition. As a result of these abnormal conditions, the refinery 150 PSIG steam header pressure was significantly low. The U205 Delayed Coker unit uses steam to purge resid and coke from the switch valve and ball valves on the coke drum structure. the low steam pressure ultimately led to the valve failing due to coke build up on the valve. On 8/28/2012 the unit was forced to go on bypass and internal circulation due to inability to switch feed to the offline drum. After the unit was on bypass the Wet Gas Compressor tripped three times. These trips resulted in releases to the flare. This resulted in a small amount of hydrocarbon material to be routed to the ground flare. Propane Flaring: Due to atypical operating conditions and the shutdown of our third-party propane pipeline, MPC flared propane starting on August 31 at 06:52 AM intermittently until September 1 at 02:45 AM. The flaring of propane was required to balance refinery operations. No reportable quantities were exceeded. The release calculations are provided in Attachment 4. North Stick Flare Flame Outage: On September 1, the North Stick Flare flame was snuffed out with steam for a total of five minutes. This occurred while decreasing the amount of propane flaring mentioned above. No reportable quantities were exceeded. This event was reported verbally on September 1st and a follow-up written report was submitted on September 7, 2012 (see Attachment 5). North Stick Flare Damage: The North Stick Flare was observed to have some abnormal flame patterns prior to Hurricane Isaac. However, during the hurricane it was noticed that one side of the flare tip had more significant flames. After the hurricane on September 11 th an inspection, via a remote helicopter, observed that a natural gas supply line to the pilots had a broken union. This was causing natural gas to burn just below the flare tip. It is believed that the high winds experienced during the hurricane caused the union to completely break apart. A repair plan is being formulated to correct this issue. Wastewater Discharge: MPC discharged untreated process area stormwater via Outfall 002 to the Lake Maurepas drainage system beginning on August 30, 2012 at 07:00 hrs intermittently until September 3, 2012 at 13:00 hrs. The amount of wastewater discharged is estimated to be 300,000 bbls (which is 12,600,000 gallons). Samples were collected prior to the discharge and after the start of discharge to verify that the water being discharged was sufficient quality to ensure no harm to environment. The discharge was monitored to ensure that there was no sheen on the water discharged off-site. It should be noted that prior to discharging the untreated process area stormwater, MPC had reached the on-site WWTP storage capacity of 619,995 bbls of water. In addition, MPC placed an out of service crude oil tank (500-2) back into service prior to the hurricane specifically to be used for wastewater and slop oil storage as needed. This tank was used for excess water storage prior to any wastewater being discharged off-site.

Followup: No

Notes: This report is linked to two LDEQ incident numbers: 142430 and 142532. Unit 205 Coker sent to the North Ground Flare. The release was identified at approximately 06:35 hours on August 29. 2012 and lasted for approximately 1944 minutes (1d 8h 24m). The compound of concern was Propylene. Totals of 76.54 lbs and 86.15 lbs were released during the 2 24-hour periods involved. MPC considers these emissions to be covered under the temporary variance issued on August 27 of 2012. That variance authorized the temporary permit for the emissions: Sulfur Dioxide 13.3 tons; Nitrogen Oxide 0.9 tons; Carbon Monoxide 7.03 tons; Volatile Organic Compounds (VOCs) 8.51 tons; Hydrogen Sulfide 0.73 tons. There was also a variance for 3,750 long tons(8,400,000lbs) of sulfur to be stored on a "sulfur pad". These variances were considered the maximum allowed during that period, therefore they are not included in LABB pollutant totals in this report. As soon as conditions allowed, the compressor was restarted. Missed Monitoring/Repair/Inspections Due to the hurricane, personnel were not available to complete several regulatory required tasks, such as weekly inspections, PMs, monitoring, and repairs. The programs and the specific missed requirements are listed below: LDAR Program 5-day first attempt at repair requirement (five components) 15-day final repair requirement (twelve components) Waste Program Weekly satellite collection area inspection (two locations) Weekly hazardous waste storage area inspection Weekly non-hazardous waste storage area inspection Weekly universal waste area inspection Weekly used oil storage area inspection Stormwater Pollution Prevention Program Weekly refinery inspection Spill Prevention Control & Countermeasures Program Weekly inspection of the Contractor Village area MACT II - NSPS Subpart UUU Weekly PM for the pH meter used for compliance demonstration Benzene Waste Operation NESHAP Carbon Canister monitoring for breakthrough (eight events) 15-day repair requirement (four sumps) There is a separate report on this database for the sinking of the "Big Tuna" response boat. Response Boat Fuel Loss: The oil spill response boat became submerged in the Mississippi River on August 29th, due to a surge in the river level which resulted in the loss of ten gallons of gasoline. The reportable quantity for oil was exceeded. This event was reported verbally on August 31 and a written follow-up report was submitted on September 7, 2012 (See Attachment 3). (Attachment 3 was deleted from this file, and added to the report for this event)
193.9 pounds

Unit 59 North Flare
Cause: A root cause analysis is being conducted to determine why this incident occurred. The Louisiana Refining Division's Investigative Summary Report states that the initiating event was the troubleshooting common alarm on the Bently Nevada radial vibration proximitor card to include a module self-test for diagnostic information.

Followup: Yes

Notes: Operations restarted the Wet Gas Compressor. The root cause analysis will result in recommendation items designed to prevent the recurrence of this event. The refinery stated that lessons learned included: On a Bently Nevada 3500 system, a module self-test or replacement will cause the output signal to go to 0ma. If the output signal is wired to a Triconex system, it will report the transmitter as being an unreliable signal (bad transmitter of bad pv). VOC chemical breakdown is provided. Only the RQ for SO2 was exceeded during this release.
7.8 pounds

Unit 259 North Ground Flare
Cause: Unit 222 Sats Gas Plant experienced an upset due to an over-pressure condition on the debutanizer column. As a result, pressure relief valves lifted, sending debutanizer overhead material to the refinery's North Ground Flare. Root cause was a faulty pressure transmitter.

Followup: No

Notes: The Board Operator decreased pressure on the debutanizer column. The faulty transmitter was replaced. Initial reports from January, 2013 stated no permit limits were exceeded concerning released pollutants. Final calculations from April 4, 2013 determined otherwise.
1.3 pounds

Unit 25 FCCU wet gas compressor shutdown
Cause: The wet gas compressor (WGC) suction flow and discharge pressure dropped suddenly, causing the WGC spillback valve to open 100%. The fractionator overhead pressure increased when the WGC spillback opened up. The high fractionator pressure SIS trip point was reached (36 psi), which tripped the unit. Fuel gas was routed to the fractionator overhead accumulator, which was being vented to the flare to keep pressure on the reactor to prevent O2 from the regen from backing into the reactor. No offsite impacts were observed by the air monitoring team. The reportable quantities for Sulfur Dioxide, HRVOCs and VOCs were exceeded. Update: cooling coil in Alkyl Unit Vent Gas Absorber (27-1107) failed.

Followup: Yes

Notes: The SIS system reacted as designed to shutdown the FCCU due to the opening of the compressor spillback valve. An incident investigation will result in recommendation items designed to prevent recurrence of this event. Update: The root causes were identified as 1) cooling coil in Alkyl Unit Vent Gas Absorber (27-1107) failed. Cause of failure is unknown. Root Cause #1: Cannot be determined until the Alkyl Unit Shutdown. Recommendation: Inspect the cooling coil in the Alkyl Unit Vent Gas Absorber (27-1107) and determine cause of failure. Based on the cause of failure, recommendations will be generated to prevent recurrence. [Complete by December 15, 2016] 2) Quaterly PMs on cooling coil in Alkyl Vent Gas Absorber failed to identify the coil was leaking. Root Cause #1: No Procedure. Recommendations: Create Operations procedure for performing the quarterly leak testing on the cooling coil in the alkyl Vent Gas Absorber. Include a step that requires operators to verify proper documentation of test result in PM work order closure. [Complete by November 18, 2014]. Root Cause #2 Preventive/Predictive Maintenance Needs Improvement. Indetify flouride sample locations for discovering a leak in Alkyl Vent Gas Absorber cooling coil [Complete by November 18, 2014].
246.0 pounds


Railcar Inlet Valve
Cause: After loading a propylene railcar, the operator noticed that the inlet valve was leaking and could not be closed. The railcar was depressured to the flare so it could be disconnected and the valve repaired.

Followup: No

Notes: None. The contents of the railcar were vented to the North Flare Stack. The railcar was sent for repairs.
70.0 pounds