Bridger Pipeline Yellowstone River Response GLENDIVE DAWSON COUNTY

Bridger Pipeline, Yellowstone River Response GLENDIVE, DAWSON COUNTY, MONTANA WINTER 2015

Initial Timeline On Saturday, January 17, 2015 Bridger Pipeline records a rapid pressure drop in its injection pipeline near Glendive, Montana at approximately 1200 Bridger immediately begins to shut down the pipeline (the procedure takes roughly 45 minutes) Bridger makes notification to NRC, Montana DES/DEQ, Dawson County DES, and the City of Glendive Initial survey crews do not see any evidence of an oil spill on the ground or in the water

Site Conditions Bloody Cold Covered in Ice

More Timeline By 2100 on January 17, Bridger has determined that the release has occurred between two block valves approximately 6800 feet apart that straddle the Yellowstone River Metering indicates that at least 300 bbls. Of oil (Bakken Crude) has been lost Calculations indicate that the pipeline volume between the valves would be an additional 925 bbls. No oil has been seen, smelled, or detected on the water or in the area

Site Location

Area Map

Sunday, January 18, 2015 OSC Peronard departs from a spill near Greeley, CO en route to Glendive, MT at 0800 2 START Contractors depart Denver, CO (with sampling/monitoring gear) en route to Glendive, MT at 0830 During the Day the City of Glendive Water Treatment Plant begins receiving odor and taste complaints from City water users OSC Peronard arrives in Glendive and tours the scene at 1700 No oil scene WTP looks normal, smells like chlorine At 1900 an Organizational Unified Command (UC) meeting is held

Priorities/Operational Theaters Safety Water Treatment and Distribution System Ice conditions, Cold, Fast Moving Water Assess, fix as needed Secure Pipeline Remove remaining oil Recover Oil From Water Assess Impacts

Safety Issues Temperature fluctuated throughout response, created unstable ice Some concern about VOCs in air Recovery of someone lost under the ice unlikely Bad form to kill someone just to recover oil

Water System On January 18, 2015, Emergency Samples collected by the MDEQ were sent to lab in Billings, MT for rush analysis. DW quality data available in two days START conducted head-space analysis (using multi-raes)of water samples at intake, in treatment plant, in distribution system, and from homes/businesses. All samples indicate the presence of VOCs in water. Up to 50 ppm VOCs detected in head space Decision to clean Water System made by UC based on head space analysis made around 1200 on Sunday, January 18, 2015, after consultation with ATSDR A do not consume advisory issued by UC around 1400 OSC Peronard ordered up mobile lab (with 2 chemists) and 2 more OSCs

Water System Plan Clean of Water Treatment Plant Flush Distribution System Flush Homes Sample and Confirm all clear City Engineer and 2 MDEQ DW Specialists to oversee In the Meantime Provide Bottled Water

Bottled Water Begun at 1800 on Monday, January 18, 2015 Secured and brought to Glendive by MDES Distribution handled by Dawson County Health Department with labor assistance provided by Bridger Issues ATSDR Opinion no acute health threat, only chronic But hard data not available for few days, I suspect levels higher Hospital State Prison

Water Plant Clean-Up Began around 1300 on Monday January 19, 2015, took about 24 hours Aerators rigged-up in pre sedimentation ponds, clear-well Dual media filters (typically only run with sand in winter, as was the case now) were subject to a “hyper “ backwash Air monitoring indicated spike in VOCs then steady drop off Carbon added back to filters (modeled based on input of 100 ppm VOCs) to remove organics

More Water Plant Clean-up Mobile Lab arrived on-site around 0130 on Tuesday, January 20, 2015. They start running samples at 0600 Mark Mc. Daniels and Scott Walker kick much ass Data indicates that the air-stripping is quite effective Mobile lab reports benzene up to 40 ppb (MCL=5 ppb) and toluene and xylene up to 700 ppb. In all 6 -12 VOCs identified in each sample Issues While mobile lab can meet DQOs/MDLs their method is not DW certified Aerators muck up turbidity. Plant intact running at around 50, 000 NTU

Distribution System Flush Started on Tuesday, January 13, 2015, around 1400. It took roughly 36 hours. Planned by the System Design/Operating Engineer, using a system distribution pressure model Worked basically upstream to downstream. Section isolated, drained and clean water pushed out from WTP No pressure loss allowed to minimize bacteriological issues VOC monitoring/head-space analysis done continuously Once monitoring showed ND for VOCs sample sent to mobile lab for confirmation (typically 2 -hour turn around)

More Distribution System Flush Staggered samples taken from WTP and Distribution System and sent to off-site lab for DW analysis, including Bac-T, once on-site lab indicates ND for VOCs Mobile lab data quality and detection levels meet or beat off-site lab Issues Concern of discharge permit (40 CFR 122. 3(d)) “Compliance Question” (NCP @ 40 CFR 300. 322(b) & (c)) Prison Unrest Surgery Cancelations

Home/Business Flush A Fact Sheet on flushing your home/business was developed by UC and City Engineer It gives a procedure based on size of home and size of water tanks Public Meeting held Thursday, January 22, 2015, to explain procedure and operations to date Call-in number and e-mail address provided for folks who need individual help Samples collected from roughly 50 homes/businesses, monitoring done at many more

All Clear for Water System Given by UC with concurrence from MDEQ on Friday, January 23, 2015, around 1300 Issues Still oil in the river, we expected a second slug with ice break-up (continued sampling at intake, and developed an intake procedure) Installation of a real-time VOC water monitoring unit at Glendive WTP not completed until February 20, 2015 With break-up (March 10 -12, 2015) a second slug did indeed come (VOC monitor worked, aerators put back on-line for 2 days, system operated at full capacity after a 12 -hour slow down)

Pipeline Recovery Operations Pipeline shut-in at block valves Pipeline unearthed and “tapped” then “cut and blanked” on both sides of the River Pipeline cleared in three steps Gravity drained Vacuumed from both ends Pressure “pigged’ from upstream to downstream 496 bbls of oil recovered from pipeline

Oil Recovery from the Yellowstone River Three Basic Techniques Attempted Ice slotting (didn’t work) Open water boom drags (not feasible as it turns out) Auger oil ice mining with sump and squeegee recovery (finally something) Recovered 60 bbls. Only 1 -2 bbls recovered after February 1, 2015 Put on hold in mid-February Oil Recovery past point of diminishing returns Ice conditions are way unstable and unsafe

Ice Slotting

Ice Slotting

Auger Oil Ice Mining Oil turned out to be trapped between layers of ice and in fissures

Oil Lost/Recovery Tally Oil Involved Lost During Pumping: Volume in Pipe: 300 bbls 925 bbls Total: 1225 bbls Oil Recovered Oil in River From Pipeline 496 bbls Discharged; 729 bbls From River 60 bbls Less recovered -60 bbls Total: 556 bbls Total: 669 bbls

Assess Impacts Good News Bad News No oil fouled flora/fauna 669 bbls lost Banks clear Some initial fish uptake No impacts to Eagles Soluble/entrained VOCs detected as far down river as Williston, ND

For More Info: Contact: Paul, 303 312 -6808; peronard. paul@epa. gov Marty, 303 312 -6963; mccomb. martin@epa. gov Joyel, 303 312 -6647; dhieux. Joyel@epa. gov Epaosc. org Click on Websites Click on Region 8 Click on Bridger Pipeline