The spill estimate for the Belle Fourche line that leaked into a small tributary of the Little Missouri last year has been revised upwards, as the company continues to investigate and uncover the causes of the Dec. 5 leak.
The company’s new estimate puts the spill at 12,615 barrels, or 529,839 gallons of oil, putting it among the state’s largest active water spills, according to North Dakota’s spill investigation program manager Bill Suess. The state’s largest spill was an 840,000 gallon spill in a wheat field near Tioga in 2013. That spill is still being cleaned up, although Suess said it looks like that could be completed this year.
The new figure for the Belle Fourche spill doesn’t yet include the hillside area, which is still leaking some oil. Belle Fourche is estimating the hillside is perhaps 20 percent of the total spill volume.
That leak has been contained, Suess said, but the hillside has been too unstable to get into for further investigation.
“The snow is drying off, so we are hoping they will have access to that soon,” he added.
Wendy Owen, a spokeswoman for True Companies, which owns the Belle Fourche Pipeline Company and the Bridger Pipeline, which had a 2015 leak in Montana that got into the Yellowstone, said the company’s initial estimates were based primarily on visual data. Since then, however, they’ve been able to get into flow rates and other verified data to revise the estimate, and as a result of their investigation there is now also a clearer picture of what caused the rupture in the first place.
Seuss said the company had shut the pipeline down Nov. 30 for routine maintenance. In that same timeframe, there was a heavy, wet snow and, without the pressure from oil in the line, the hillside slumped, bending the pipeline in two places, including at a weld, where it split open.
“That would explain why the leak detection didn’t work,” Suess said. “(The electronic monitoring system) is gonna go take a baseline, but if the baseline is a leak, it’s not going to catch it.”
The company has since implemented other strategies, Suess said, to improve their methods for catching that type of scenario.
“The technology on gathering systems is always tricky,” Suess said. “Unlike transmission lines, where there is a consistent volume from point a to b, with gathering lines you have have multiple points. There might be a point A, a point B, C and D, and all those different areas can be on or offline, so it’s hard to measure the volume going through them because it’s not consistent.”
However, pressure data can help discern that there is a problem with the line of this nature.
“If you look at pressure flows along the line and move down and suddenly it’s lower, then you know you have a problem,” Suess said. “Or if you’re starting up and your pressure is not as high as it should be.”
Suess says it is common for initial estimates to be less than actual amounts, and for there to be some ambiguity as to what exactly happened. The initial focus is on containment and cleanup, not on investigating the total spill amount and its causes.
“When it was first released, they were thinking it was a day, then they got in, looking at the records and numbers,” he said. “It takes time to do that. Plus the fact the creek is frozen when it first occurs, so it’s hard to get a handle on the volume. It’s understandable that they are low.”
Owen said the company is probably 80 percent of the way finished with its cleanup. They are focusing right now on the Ash Creek Coulee, which is a water source for cattle. Grazing season begins May 1, and the hope is they can be finished with cleanup so that the area can be opened by then.
Owen said they will be on the site until cleanup is completely finished, no matter how long it takes.
After they are done with the coulee, efforts will shift to the hillside, which could be a more lengthy process, Suess said. The company will explore whether an in situ bioremediation, which uses oil-eating bacteria, will be effective. That usually takes longer than excavation, but disturbs soil less, and leaves behind organic material that studies have shown to be beneficial. Crude oil is an organic substance, and there are naturally occurring bacteria that will consume it, turning its longer chain hydrocarbons into benign, organic material. In that sense, oil spills are sometimes easier to address than briny, produced water spills.
Along the coulee, the company was able to use burning to speed its cleanup efforts along, Suess said. They did 1,200 burns, which he described as very effective.
“In situ burns are not new,” Suess added, “but it’s the first time they’ve really done it on the water in North Dakota. They did one down in the gulf with the BP spill. So they’ve done them before, but this is the first in North Dakota.”
Suess said the technique works well with Bakken oil spills on water, and could be used again in the state if there is another such spill.
“Hopefully we don’t have too many more like this, though,” he said.
While the North Dakota Department of Health has issued a notice of violation for the spill because it escaped containment and it affected water, they have not yet calculated an amount for the fine.
“We wait until well into cleanup for that,” Suess said. “How well they do with the cleanup, how aggressive they are plays a part in how much the penalty is.”
Other factors Suess said will be taken into account include how long before the spill was contained, and how long the company took to clean it up, as well as the total spill volume, once the investigation is complete.
Suess said the cleanup effort has been going well, and that testing on the Little Missouri has so far not found benzenes at levels of concern.
A warm stretch in February did cause some flash flooding, which overtopped the containment berms at the site, but the dilution factor from so much sudden water, plus the volatility of Bakken crude, helped keep the levels of contaminants below thresholds of concern. The company has since replaced and beefed up the containment.