One million barrels is a lot of crude oil—42 million gallons, 159 million liters—enough to fill 63-and-a-half Olympicsize swimming pools. Now, imagine having to move that volume of liquid every day to someplace, somehow.

That’s the job Williston basin midstream operators face in the Bakken, the sprawling unconventional shale play that laps across a big slice of western North Dakota, slides under the state line into Montana and goes northward into Saskatchewan and Manitoba.

Bakken production had been ramping up steadily in the past decade, month after month, year after year. Multiple
industry observers predicted several years ago the Bakken, given its big geologic footprint and hydrocarbons in place, might-maybe-possibly-somehow-could be producing 1 million barrels (bbl.) per day of crude at some point off in the future.

That point has come and gone.

The U.S. Energy Information Administration (EIA) estimates the Bakken clicked past the million-barrel mark in December 2013. January production rose to an estimated 1.011 million bbl. per day and EIA’s February estimate rose again to 1.036 million bbl. per day, although bad weather may have cut into that number. So the Bakken is among the world’s biggest oil fields.

To benchmark the Bakken’s rapid growth, remember that five years ago its output was around 100,000 bbl. per day.

And gas, too

Associated natural gas production numbers impress as well, with EIA setting Bakken gas output at 1.12 billion cubic feet (Bcf) per day in January, rising to an estimated 1.14 Bcf in February. That’s comparatively small compared to the Marcellus (14 Bcf per day) or the Eagle Ford (6.2 Bcf per day)—but still significant for a play that just blipped on the energy industry’s radar for the first time less than a decade back.

If one considers the development and sheer volume of crude and gas production coming from the Bakken/Three Forks Play, then clearly the Williston basin must have a huge impact on infrastructure throughout the U.S. and Canada. Pipelines for crude and natural gas are continuing to expand, while rail is still a major contender for transporting crude to various destinations around the country.

Bear in mind the Bakken also represents a major Canadian unconventional play, Foster Mellen, senior strategic analyst for Ernst & Young’s EY Oil & Gas, tells Midstream Business.

“From our understanding and work with our Canadian practitioners, things are ramping up quite noticeably on the Saskatchewan side,” he says. “There, the producers are very, very optimistic that this is going to be a major production basin for Canada.”

Although not in the same league as western Canada’s oil sands, “certainly it’s nowhere near the scale we’re seeing in Alberta or British Columbia,” he adds, “nonetheless, a big deal.”

The Bakken and the Three Forks offer an oily version of the unconventional shale plays that have created multiple midstream opportunities across North America—plus a fair amount of associated gas and natural gas liquids (NGLs), as noted above.

The play challenges upstream exploration and production firms that find and produce all those hydrocarbons.

The formations have some comparatively complex geology. Generic references to “the Bakken” typically refer to a stacked, multi-play Bakken formation, which has separate upper, middle and lower producing zones. Also, the geologically similar Three Forks formation lies farther down but to date has been lightly drilled. It could
become a major producer in its own right.

Late Devonian to early Mississippian in geologic age, the source rocks are tight, low-porosity shales and dolomites that resist conventional drilling and completion techniques. But natural fracturing within the formations enhance horizontal drilling and hydraulic fracturing. The first horizontal Bakken well went down in 2000 in eastern Montana.

Bakken wells typically produce a light, sweet crude of around 42º API gravity—similar to and a little lighter than the West Texas Intermediate that serves as a commodity market and refining-industry benchmark. The Bakken crude has a significant amount of heavier NGLs, which accounts for the current quandary of how rail shippers can best handle the volatile oil.

Gathering constraints

The geographically remote Bakken proves the adage about value and the midstream: Upstream producers can find lots of oil and gas, but they don’t make a cent until it moves to market.

Getting production to market has been a major challenge in a region that historically had comparatively little oil production and little local demand. North Dakota has one refinery, and its largest city has a population of about 60,000.

David Scobel, chief operating officer at Denver-based Caliber Midstream Partners LP, oversees an operation focused
on the Bakken. Just keeping up with the rapid drilling by the play’s producers represents a significant midstream challenge for the first link in the midstream chain—gathering, he says.

“We certainly see at the gathering level that there’s a ways to go on the infrastructure all the way around,” he tells Midstream Business, adding that some estimates say half of the Bakken wells drilled to date aren’t on pipe yet.

“The larger takeaway projects, those are fairly well known; the capacity constraints that the market will struggle with on takeaway. We’ve all been talking about this. There’s a lot of infrastructure that needs to go into the gathering systems to get to the wellhead,” Scobel adds.

The Bakken has its own unusual challenges to midstream operators, he points out. One is how to allow for the steep decline curves typical of Bakken wells, which typically come on at very high initial production (IP) rates, then drop steeply during the first few months of production, then stabilize at much-lower sustained rates for many years.

For example, Statoil drilled three McKenzie County, North Dakota, wells last year that had IP rates of more than 5,000 bbl. per day. Gathering lines that are big enough to handle production like that quickly become underutilized as wells drop off to perhaps a 100 bbl. per day.

Pad drilling

A second challenge has been Bakken producers’ use of pad drilling, in which multiple wells, perhaps a dozen, are drilled from one location to cut downtime and rig relocation costs.

That’s fine, Scobel says, but how does a midstream operator best respond to all that drilling from one small location? Does the midstream service provider wait until perhaps six or eight wells at one location are drilled, fraced and completed to run the necessary gathering lines out to the pad? Or, would it be better to nickel-and-dime
the pad as wells come on one at a time?

“Likely it’s going to be some type of hybrid,” Scobel speculates. “Most producers are doing at least two-well zipper fracs to optimize their completions. We think that you could see four- or six-well completions and initial productions coming off of one pad.

“So now you’ve built a 4-inch gathering line out to these pads and you have an initial (production) that’s coming on at 5,000 bbl. a day—and you’ve undersized your system before you’ve even taken a barrel. You still have a lot of trucks that are having to bring fluids to location and take fluids from the location” as a result, he says.

Gathering includes more than crude but also natural gas—and water, lots of produced water.

Bakken wells produce water with gusto and it has to go to disposal wells, another challenge for midstream service providers. Water-gathering systems must go in to reduce heavy truck traffic on county roads.

“We’re trying to change the game, and we’re putting in some larger gathering headers for produced water and for crude than would be typical at the gathering level, just because we believe that continued (well) downspacing is going
to be successful, and you are going to see higher IPs,” he says.

To market, to market

Once gathered and processed, the crude has to move to market.

EY’s Mellen says there’s considerable Bakken pipeline infrastructure going in despite all the talk about crude by rail. Hart Energy’s North American Shale Quarterly (NASQ) rates the play midway through a midstream build-out that will run through 2017.

“There’s a lot of smaller pipeline capacity being built out of the Bakken to tie in with the mainline systems, particularly Enbridge—the big, big crude pipeline—and also with the TransCanada pipeline that’s there,” Mellen says. “But again, that crude is mostly designed to move south. I suspect there will be some pipeline capacity proposed to be built to tie into the proposals to move crude farther east. I wouldn’t be surprised to see some of the Bakken crude in the Energy East pipeline, once it gets built by TransCanada.”

TransCanada’s proposed Keystone XL pipeline, viewed primarily as a transporter of Canada’s heavy oil sands crude, likely would handle a sizeable share of Bakken crude too, probably around 100,000 bbl. per day.

“We’d like to think that that's eventually going to be permitted,” Mellen says of the Keystone XL.

“The eastern Canada refineries, while some of them can handle the heavy Canadian (oil sands crude), they’re still big users of the lighter blends. So we’re seeing the proposed Line 9 Reversal to Montreal, which is the Enbridge deal. But TransCanada’s Energy East should come about, we have no reason to think it wouldn’t. It could also attract some of the lighter blends as well.”

Enbridge has a proposal to Canada’s National Energy Board to reverse the flow of its 396-mile Line 9, which runs from Montreal southwest into Ontario. It has been flowing westward but reversing its flow would expedite the flow of North American crude in the 240,000 bbl. per day line to Québec’s two refineries, which represent 20% of Canada’s total refining capacity.

The project has drawn extensive opposition from environmentalists who point to Enbridge’s 2010 oil spill in Michigan.

TransCanada’s 2,700-mile Energy East project will move as much as 1.1 million bbl. per day of crude from western Canada to eastern Canadian refineries and export terminals. But it’s likely to pick up a fair amount of Bakken crude en route across Saskatchewan.

The project is expected to cost $12 billion and will convert a 1,864-mile portion of TransCanada’s existing Canadian Mainline gas pipeline to crude service, as well as add 870 miles of new pipe.

Crude will be transported from Alberta’s oil sands to delivery points in Montreal and Québec City and terminate at Canaport in Saint John, New Brunswick. The Canaport terminal also will house a new deepwater marine terminal that TransCanada is developing with Irving Oil. The marine access would allow Canadian producers to sell crude to lucrative Asian markets, such as India, where Canada can compete with production from the Middle East and Africa.

The Pony Express

In the U.S., one of the larger, Bakken-focused pipeline projects under way is Tallgrass Energy Partners’ Pony Express Pipeline, which has received a necessary abandonment order for existing gas service from the Federal Energy Regulatory Commission. A 430-mile section will be converted to handle crude and joined to 260 miles of newly laid line to create a link between North Dakota and the major crude-trading hub at Cushing, Oklahoma.

The 24-inch line could ultimately move 320,000 bbl. per day. Completion is scheduled for the first half of 2014.

More crude at Cushing may not be Bakken producers’ first choice. The hub has suffered from a major glut as growing
Bakken, Permian and Niobrara oil flowed into its tanks. That flow, frankly, tanked prices for domestic oil relative to
the major foreign benchmark crude, North Sea Brent. However, new pipeline capacity to the south, serving Gulf Coast
refineries, seems to be easing the pricing problem.

That midstream bottleneck has had a predictable impact on crude prices at Cushing and elsewhere, although the gap has narrowed as new midstream capacity has been added.

“Bakken crude with landed cost at the Northeastern United States, for example, traded at a significant discount to Brent prices because of midstream capacity constraints,” according to a recent Deloitte report, “The Rise
of the Midstream.”

It says, “In 2011 and 2012, that discount averaged $21.50 per bbl., which was greater than the average rail cost of $15-$16 per bbl. to the Northeastern United States and U.S. Gulf Coast. New pipelines to the U.S. Gulf Coast, however, cut the Brent/Bakken discount to $11 per bbl. in early 2013, an indication that other forms of transportation may become too costly to be a long-term solution for a lack of pipeline capacity.”

Head east

An Enbridge subsidiary, North Dakota Pipeline Co. LLC (NDPC), closed an open season at the end of January for capacity on its $2.6 billion Sandpiper expansion project, which will move Bakken crude eastward. The proposed line will expand and extend the takeaway capacity of Enbridge’s existing Williston system to 580,000 bbl. per day from 225,000 bbl. per day. Marathon Petroleum Corp. signed up as anchor shipper.

The new 375-mile, 24-inch line will extend from Beaver Lodge, North Dakota—the current terminus of the system—to Clearbrook, Minnesota. The project also includes a new 375,000 bbl. per day, 233-mile, 30-inch diameter line to be built extending the NDPC system from Clearbrook to connect with the affiliated Lakehead Pipeline mainline terminal at Superior, Wisconsin. Enbridge’s expected in-service date is first-quarter 2016.

Rail responds

North America’s crude-by-rail phenomenon started here—first viewed as a stopgap until new pipelines went in the ground—and is now regarded as a major midstream player in its own right. In comparison to scant existing pipeline capacity, the Williston basin had a comparatively good railroad network in place, thanks to the transcontinental railroads the U.S. and Canada built across the region in the late 19th century.

Those systems still offer excellent east-west links to the rail hub at Chicago and Atlantic refining centers and to big Pacific ports such as Portland, Seattle and Vancouver. Southbound rail service is not as well developed but also
not as much in demand.

With that great rail infrastructure, estimates place nearly three-quarters of daily Bakken production moving to
market via tracks rather than pipes. (See Whither Rail?)

Rail may be a permanent Bakken midstream player. Rail serves markets—particularly the Atlantic and Pacific coasts—that are unlikely to ever have a pipeline link to the Great Plains due to distance, topography, large population centers and environmental concerns.

Although new pipeline capacity to eastern Canada or additional capacity south to the U.S. Gulf Coast could happen, it’s highly unlikely that new pipeline capacity will be built to two of the largest markets for Bakken crude—the U.S. east and west coasts. Refineries on the coasts welcome attractively priced Bakken crude as an alternative to more costly North Sea Brent or Alaska North Slope crudes. Multiple new spur tracks and loading terminals have gone in alongside the major rail lines in the Bakken.

The problem is that rail terminals also must be available at each end of a railroad line. Mellen says that has been a challenge—particularly on the West Coast.

“Permitting is hard in California for everything,” he says. “It’s not just rail, and it’s not just oil and gas. The other rail terminal projects that have been going on, while I’m sure there was a bit of local opposition at some point, nothing really rose to the surface as far as my understanding.”

One major project boosting rail capacity has been Dakota Plains Holdings’ $50 million Pioneer project, a substantial expansion of its existing terminal operation at New Town, North Dakota. Completed at year-end 2013, the 192-acre site has twin 8,300-ft. loop tracks each capable of handling 120-car unit trains, a 10-car loadout building, two 90,000-bbl. storage tanks, a 10-station truck rack and five pipeline interconnections.

Meanwhile, an inbound oilfield products business at the Pioneer terminal continues to develop with the first frac sand transloading that was scheduled to start earlier this year. Completion of the inbound side of the terminal is scheduled for May.

This expansion allows the terminal’s four, existing 2,500-foot tracks to be used for inbound oilfield commodity supplies, such as frac sand. Some 90 acres has been set aside for future industrial development.

Capital constraints

All of these midstream projects take money to complete, so will the capital be there?

Mellen answers “yes and no,” then explains that conflicting answer.

Capital is not constrained at the moment, but it’s a long way from free flowing, he says. “I think the biggest threat in North Dakota and the Bakken would be a sustained dip in oil prices. Bakken production, as a whole, is at the high end of the cost curve. If we had global oil prices drop down in the $70 to $80 bbl. range, I think that would put a big hurt on a lot of the producers.

“So not only would we see growth slowing down, but also we would probably start to see declines in production—
just because of the decline curves up there, particularly the small to mid-sized companies strapped for reinvestment dollars. Right now there seems to be enough capital for people to do their expansions, but it could tip fairly quickly.”

The Deloitte report elaborated on that point.

“Despite this rise in its capital intensity, the [midstream] sector is just beginning to meet the needs of exploration and production companies in areas like North Dakota’s Bakken shale formation,” Deloitte said, pointing to the sizeable associated gas output. “According to the latest available annual data (2011) from the EIA, about 35% of the Bakken natural gas production had to be flared or was otherwise not marketed because of the insufficiency in the infrastructure required to store or transport it.”

The gas problem

headache for oil-focused producers and midstream operators alike. The Bakken’s gas production is sizeable but spread across thousands of wells that are spaced out on thousands of square-mile land sections. It takes a lot of infrastructure—gathering lines, processing plants and transmission lines—to move all of it out.

The result has been significant gas flaring. Nobody likes it. Even some royalty owners have grumbled that flared gas reduces their sizeable royalty checks. Flared gas is lost revenue and creates environmental problems, but gas has had
to take its place in line behind more profitable crude.

The North Dakota Petroleum Council recently formed a flaring task force that has proposed flaring be limited to around 5% of total gas production. Opportunity creates interest, and Bakken gas is a bona fide opportunity for midstream players.

Gas-focused ONEOK Partners has expanded its Bakken gathering and processing operation to handle the underserved Bakken gas business. The company recently told investment analysts it has some $2.5 billion in capital projects planned for the Williston basin.

ONEOK has four Bakken gas plants in service now with a combined capacity of more than 300 million cubic feet (MMcf) per day. The 100 MMcf per day Garden Creek II plant is scheduled to go on stream in the third quarter while two more plants, Garden Creek II (100 MMcf per day) and Lonesome Creek (200 MMcf per day), are expected to enter service in 2015.

Gas liquids can move south via ONEOK’s Bakken NGL Pipeline to the Conway, Kansas, NGL hub and eventually to the Gulf Coast and its numerous refineries and petrochemical plants.

The line currently has a capacity of 60,000 bbl. per day but a $100 million expansion will raise capacity to 135,000 bbl. per day by the third quarter. A second expansion to 160,000 bbl. per day could happen in the first half of 2016.

ONEOK can move the residue gas from its plants to its 50%-owned trunkline Northern Border system or Viking Gas Transmission.

Bernie Colson, managing director of energy infrastructure equity research for Oppenheimer & Co., sees significant value in ONEOK’s Bakken assets.

“Continued Bakken production growth is expected to drive attractive volume growth across [ONEOK Partners’] assets,” he said in a recent report. “Current infrastructure forms a highly strategic system that positions [it] to continue playing a key role providing services to the rapidly expanding petrochemical industry.”

The flare over there

Scobel says Caliber and its peers are working overtime on the gas problem.

“I think the midstream folks are going to struggle getting large-enough pipe and compression to the wellhead to minimize flaring,” he says. “One of the alternatives is, rather than try to get all the volumes 100 miles back to a central plant, put more smaller plants out in the field level. And we think that's one way that we can be effective there.

“But there's a type of flaring that nobody is talking about right now, and we’re in the middle of it,” Scobel adds. “If you were to connect every well right now that’s producing—and we’re at 10,000 wells—to pipe and you had enough pipe compression and plant capacity to take all that associated gas, the basin would still be flaring between 50 MMcf and 100 MMcf per day of tank vapors, and that’s not really being addressed anywhere.

“But we’re taking that head on; we have a method where we can take high-vapor-pressure crude with the gas entrained in the crude that would be normally flaring off at the tanks. We bypass the tanks altogether. We take crude directly from the separator into our pipeline and we have tanks there only for upset, we wouldn’t eliminate them,” he adds.

Pick a number

That million-a-day milepost just flew by. So how high will Bakken production go?

Hart Energy’s NASQ projected in its fourth-quarter 2013 report that Bakken production will peak at around 1.4 million bbl. per day in 2022 and slowly decline from that point.

There are a lot of other big production numbers floating around, Scobel says.

“I think on the conservative side you see numbers that say we’re going to be at least 1.4 million bbl. a day as a peak,” he adds. “We think it’s probably closer to 2 million without being too aggressive. So, yeah, we don’t see any end in sight. And again, the [midstream] infrastructure is going to have to match that volume, right?”

Paul Hart can be reached at pdhart@hartenergy.com or 713-260-6427.