“We have to take seriously the volume of crude that the U.S. has got to export,” Dan Lippe, principal of Petral Consulting, told Midstream Business. “Predominantly that is now light sweet crudes, but over time it will be whatever the domestic refining sector does not want.”

It is simple economics, Chris Hedge, a director in the process and technology practice at consultancy Opportune, told Midstream Business. “We have more supply in the U.S. than we have demand. Without access to more demand, growth is capped. Exports provide access to more demand and supply can grow. Add to this the fact that U.S. supply is low-cost compared with other areas of supply, and U.S. barrels will compete nicely in the global markets.”

Not only are crude exports now necessary, the industry already saw the same thing happen on a smaller scale a few years ago.

“This has already happened in propane and NGL,” said Greg Haas, director of integrated energy for Stratas Advisors, a Hart Energy company. “If there had not been export approvals for NGL and sufficient pipelines and loading facilities for them as well, we would not have seen the shale gale that we did see,” Haas told Midstream Business. “Millions of barrels of liquids cannot just go into storage without affecting domestic prices and production and, by extension, the international markets.”

Exact numbers vary, but there is consensus that U.S. crude exports have been rising steadily since the 40-year ban on most exports was lifted in December 2015. Crude exports averaged 590,000 barrels per day (Mbbl/d) in 2016, 1.1 million barrels per day (MMbbl/d) in 2017, and more than 1.9 MMbbl/d so far in 2018.

Exports hit an all-time high of 3 MMbbl/d in June.

“Those numbers are fairly in line,” John Coleman, senior analyst for North American crude markets at Wood Mackenzie, told Midstream Business. “It is important to note the 3 MMbbl/d number was one weekly estimate by the U.S. Energy Information Administration (EIA) and subject to cargo timings. I should note the monthly average for June was closer to 2 MMbbl/d. Our outlook going forward is for U.S. exports to continue growing and to more than double before reaching a peak in the 2030s at more than 4.5 MMbbl/d.”

According to Opportune, U.S. crude exports have grown 5% per month on average compound monthly growth rate since January 2016, and have increased to 8.1% from January through August 2018.

“If—and it’s a big if—that trend were to continue,” Hedge said, “It is feasible that we could hit the 3 MMbbl/d monthly average mark by year-end 2018. That would mean loading one more very large crude carrier (VLCC) at 2 MMbbl about every 2.5 days, or approximately 12 more per month. The other option would be 800 Mbbl/d moved through pipelines to Canada or Mexico, or some combination of those options. Based upon my assessment, a sustainable rate of about 2.4 MMbbl/d by year end is more realistic.”

Refining, rising

Haas at Stratas is sanguine about consumption rising domestically, which will take some of the edge off the heavy export needs both volumetrically and economically.

“Domestic crudes are trading at a substantial discount to Brent,” he noted, “and domestic gas is trading at a substantial discount to global benchmarks as well. U.S. refined products are at a premium, which explains high exports and record U.S. refinery operating rates. It makes all the sense in the world to refine discounted crude to make fuels that sell on the international markets at Brent-based prices.”

Lippe at Petral noted that “in just a few years the two biggest refineries in Corpus Christi [Texas] had their crude towers rebuilt to be able to process the increasing supply of light sweet crude from the Eagle Ford. Within three to five years, more majors will retrofit some of their large refineries.”

The issue for U.S. refiners is more around crude quality and the light sweet crude being produced in the U.S., Coleman said. “With U.S. refiners close to being maxed out on light sweet crude, it is true that the global market is now required to balance U.S. production. For U.S. production to continue to grow, export markets will be required to absorb nearly all production growth.”

Current U.S. refining capacity is 18.6 MMbbl/d, according to the American Fuel & Petrochemical Manufacturers trade association. Of that, 45%, or 8.4 MMbbl/d, are around the Gulf Coast. Stratas expects that something close to an additional 1 MMbbl/d of additional refining capacity can be expected in the next several years.

“ExxonMobil has been talking about increasing crude capacity at its refinery at Beaumont, Texas, since its first-quarter earnings call,” said Haas. “That could mean adding something like 400 Mbbl/d to 600 Mbbl/d of crude capacity, which could make the facility the largest refinery in North America.”

ExxonMobil’s Beaumont, Texas, refinery has a current capacity of about 360 Mbbl/d. The largest in North America is just down the road, at Port Arthur, Texas. That Motiva-owned monster can process 635 Mbbl/d.

In a small irony, the next-largest tranche of additional capacity may come from the facility that once was the largest refinery serving North America. At one time the Hovensa complex at Limetree Bay on St. Croix, U.S. Virgin Islands (USVI), was No. 1 at 650 Mbbl/d—bigger than Motiva Port Arthur today. It was a joint venture of Hess Corp. and Venezuela’s PDVSA but was shuttered in 2012. It has since been owned by an ArcLight Capital Partners portfolio company that has operated it as a terminal.

“That is likely to be restarted with a crude consumption of about 200 Mbbl/d to 300 Mbbl/d,” Haas said. “Even without any other major announcements of refinery expansions, we anticipate incremental creep of 10 Mbbl/d, 20 Mbbl/d or 50 Mbbl/d at multiple facilities.

“There are also the topping plants that have already popped up in a few shale basins and are likely to continue being built as boutique refineries to meet local demand. All of that could be as much as 200 Mbbl/d. Those, with Hovensa and Exxon, easily get to 1 MMbbl of expanded North American crude refining capacity,” Haas added.

Haas also noted a singular attribute at Hovensa, highly relevant to export markets.

No Jones Act
“The U.S. Virgin Islands are, of course, U.S. territory, but they are specifically exempt from the Jones Act.” Formally the Merchant Marine Act of 1920, it stipulates that trade between U.S. ports can only be conducted by vessels built in the U.S., owned by U.S. companies, and with an American crew. Many countries have similar cabotage rules, but the dearth of compliant vessels for the coastwise trade makes the Jones Act expensive for U.S. shippers.

“The exemption means that vessels of any nation can take U.S. crude from U.S. ports to the refinery at St. Croix,” Haas said. “The same is true of fuels going from a restarted USVI refinery perhaps back to the mainland U.S. This exemption will save several dollars per barrel on crude and on refined products.”

The exemption is not a secret. Indeed, BP Plc is thought to be a possible partner in the Hovensa restart. That would make sense, in the context of the company increasing its activity in the Gulf of Mexico. Despite the Macondo disaster in 2010 that killed 11 people, BP still operates four platforms in the Gulf: Thunder Horse, Atlantis, Mad Dog and Na Kika, and has two other large projects underway.

The closest markets
Exports to neighboring nations should not be ignored; Haas noted that Mexico is interested in buying some volumes, most recently of Louisiana Light Sweet grade, and also that Canada imports significant volumes of condensate. Most of that is used as diluent for bitumen. The resulting “dilbit”’ blend flows easily through pipelines, or in and out of general-purpose rail tank cars. In contrast, “railbit” is almost straight bitumen and relies on specially heated railcars for flow viscosity. It also requires heated loading and unloading terminals.

Canada has had its own vexations getting crude either to the U.S. or to tidewater. Physically getting crude from Alberta over the Rocky Mountains is nothing compared with the political difficulty of getting through the province of British Columbia and then to the ocean.

“We are not taking Keystone XL [from Canada to the U.S.] or the Trans Mountain expansion [through British Columbia] as done deals,” Haas said cautiously. “The deal that is as done as can be is the Line 3 expansion by Enbridge [from Alberta to Superior, Wis., on the Great Lakes]. That is expected to provide about 375 Mbbl/d of capacity in the latter half of 2019. There is another 450 Mbbl/d of system-wide optimization that is possible; with Line 3 that adds up to 825 Mbbl/d by 2020. That is effectively equal to Keystone XL.”

Similarly, Mexico’s market may be more of a swap than a purchase. The bulk of the country’s production is heavy, but its old and inefficient refineries are geared to light crudes. Several sources suggest that Pemex would be wise to invest its limited capital in new production rather than trying to upgrade its refineries. Imports of light crudes from the U.S. would be used to optimize the feedslates for those facilities.

Assessing the need

Commenting on the rush of export terminal plans—there are at least four and as many as seven—Haas said it is similar to the rush to build rail-transfer stations in the early days of the shale bonanza, before pipelines could be built.

“The urgency has taken the industry by storm,” he said. “At one point, developers in Canada started more than 800 Mbbl/d of crude-by-rail loading capacity. But actual loadings were 174 Mbbl/d before the 2014 oil price crash and fell afterward to as low as 50 Mbbl/d, and have since recovered to four times that. But I doubt we are ever going to see that 800 Mbbl level.”

Stratas calculated that as of June, marine terminal capacity for loading crude along the Gulf Coast totaled 4.5 MMbbl/d. But on that list, only one, the Louisiana Offshore Oil Platform (LOOP) can fully load VLCCs without more costly reverse lightering. In reverse lightering, the big tankers take on a partial load at shoreline docks, allowing them to navigate comparatively shallow channels. Once at sea, other vessels top off their loads before they sail to customers.

By 2020, the U.S. may have as much as 7.5 MMbbl/d of loading capacity if all the terminal projects proceed as expected, Stratas calculates.

Haas recognizes that is a big “if.”

“Our forecasts of available crude for export from the U.S. Gulf Coast do not support a need for 7.5 MMbbl/d of export capacity. It’s more like half or two-thirds of that at best. That’s why this reminds us of the crude-by-rail terminal loading rush earlier in the decade that saw numerous smaller manifest loading terminals replaced by unit-train facilities, neither of which are being fully utilized now,” Haas said.

While definite capacities for the proposed VLCC terminals for the Gulf Coast are not yet set, Haas said that at an average of 750 Mbbl/d, there could easily be a need for at least two or three.

“We could easily accommodate four or five large terminals, but those would start to cannibalize the loading volumes from smaller docks,” he added.

Apples and oranges
Capacity assessments with regards to marine movements are always a little tricky because they are not ratable the way a pipeline is, noted Hedge at Opportune.

“While one of these facilities can load at 60,000 barrels per hour, that doesn’t mean it can export 1.4 MMbbl/d every day. The ships take time to get into position and get connected, then be loaded and disconnected, then maneuver away. All that adds one to three days to the cycle time and takes the actual capacity down to one-third of the actual pump rate,” he said.

It is also important to note that tanker operators do not want to pay demurrage costs, so ships are not going to be waiting in line to be loaded. Thus there can be days between vessels.

“From a terminal owner perspective,” Hedge said, “they want the U.S. to have just enough export capacity to not induce someone to build another terminal. From a shipper’s perspective, they want as much capacity available as there can be. Competition keeps the cost down.” Multiple facilities also mean a more reliable supply chain overall, and possibly alternatives for shippers if there are disruptions.

According to Opportune, if U.S. crude production were to increase from 11 MMbbl/d to just 11.5 MMbbl/d—an increase of merely 4.5%—that equates to more than seven additional VLCC cargoes a month.

“In that case,” Hedge said, “assuming all additional barrels could flow to the facilities equally without constraint, and that increased production is not offset by a decrease in imports or increase in refining demand, then just one new terminal would suffice. Two would be better. Three would probably be one too many.”

How many racehorses?

The one VLCC-capable loading facility, the LOOP off Port Fourchon, La., has 72 MMbbl of storage at Clovelly, La., but only limited pipeline capacity to major Texas and Oklahoma shale plays.

LOOP was originally built as an import terminal but was reconfigured to allow bidirectional flow and load for export.

“While I have not heard of any new pipelines in consideration to Clovelly, it is my understanding that the LOCAP and Capline pipelines are being considered for reversal,” Hedge said. “This will be a key step in making the LOOP an essential export player. LOCAP connects Clovelly to the St James [La.] terminal and by extension to crudes coming in from the west and north, while Capline currently moves crude from Clovelly north into the Midwest refining region. A reversal of Capline could bring heavy Canadian and Bakken crudes into Clovelly.”

Coleman at WoodMac explained that “the broader theme will be getting VLCC loading access for Texas and Oklahoma producers where there is current pipeline connectivity off the Texas coast. Several offshore and onshore VLCC-capable projects have been put forward to provide this access. The LOOP export terminal will likely specialize in Gulf of Mexico medium-grade exports, which have been seen in its limited export operation to date.”

Looking more closely at some of those proposed loading facilities, Coleman added, “We expect the Tallgrass/Drexel-Hamilton Seahorse Pipeline from Cushing, Okla., to St. James, La., to move forward. A key selling point may be the ‘stem-to-stern’ service that Tallgrass can offer Rockies producers by transporting crude from the wellhead to export dock on one integrated system.”

Similarly, the Trafigura offshore terminal at Corpus Christi “is likely in our view and moving forward through permitting currently. Also, the Enterprise [Products Partners] project is very real. It is likely this project will move forward. Enterprise already has the most extensive crude infrastructure and aggregating capability in the U.S. Gulf Coast market,” Coleman said.

Conversely, Coleman reckoned that the Jupiter MLP pipeline from the Permian to Brownsville, Texas, “is unlikely in our opinion given the lack of current infrastructure in Brownsville.”

The overarching question in handicapping the various proposals is the total need for number of facilities and cumulative throughput.

Like Haas said, “There is likely overbuild in export capacity looming if all these projects move forward,” according to Coleman. “Our current estimate of onshore capacity, should all proposed terminal projects inland move forward is greater than 6 MMbbl/d in capacity. Offshore terminals would be substantially additive to that figure. This compares to our peak U.S. crude export figure of 4.5 MMbbl/d—implying that not all projects on the table today will be necessary, or poor return on investment might be expected for midstream operators if an over-build comes to fruition.”

From shale to shore

Any export terminal assumes that there is sufficient pipe to get the oil from wellheads to tanks for loading. That does not seem to be an unrealistic expectation, Coleman said, once the midstream sector gets past the current tightness.

“The most critical avenues for U.S. crude into the future will really begin and end with the Permian basin,” Coleman said. “That region will be the growth engine for the Lower 48 going forward. Having enough pipeline capacity to get that crude production growth into the U.S. Gulf Coast market for export will be critical for the growth to continue. While it is constrained today, huge slugs of capacity are expected online over the next two years to support Permian production well into the future.”

All the focus on marine send-out terminals assumes that the midstream sector in the U.S. and Canada will have stepped up and built enough pipe to get from the wellhead to the tanker. Haas said that is a fair expectation.

“With the Capline reversal, the eastern Gulf Coast will be well-served, even including the Diamond line bringing Texas and Midcontinent crude from Cushing to an intertie at Memphis. There are also other important projects flowing to intermediate hubs with existing or pending access to the Louisiana or Texas Gulf Coast, including the Dakota Access Pipeline,” he said.

But the Texas markets will be the biggest winners as export volumes burgeon, Haas said. “Corpus Christi is closest to the Eagle Ford and the Permian,” he added. “For crude exports, that makes it a one-drop shop and a one-hop stop. Many Texas barrels can be arranged on one contract with one tariff on the way to the export terminal.”

Still, shippers like options, and the global oil market is notoriously volatile. The safest bet is to send crude to an area where there is the alternative of large refining operations if export markets are difficult and differentials tighten to close the export window. That favors the stretch from Houston to Beaumont-Port Arthur, over to Corpus Christi or onto Louisiana. And, Haas added, “There is no refining at all at Cushing.” The Midcontinent pipeline hub’s last refinery closed in the 1980s.

In handicapping the four proposed export terminals, Lippe is not discounting any of them, but is a bit skeptical about two of them.

“I don’t see the advantage of the Tallgrass proposal. Why pump crude from Texas to Louisiana? And why pump more crude to Cushing than already goes there? The oil is coming from South Texas and West Texas and is going out through the Gulf Coast. Why go inland to Oklahoma if you don’t have to?” he asked.

That said, Lippe is also less than sanguine about the Trafigura plan for Corpus Christi.

“There is some refining there, but not a great deal, and none of the major oil producers have capacity there. Corpus is just not a primary destination. Houston, Beaumont, and Port Arthur are the largest regional concentrations of refining,” he said.

Shippers like options

Lippe said that the most prudent approach is for expanding crude production to be sent where export or processing are equal options. That gives shippers, refiners and traders the widest diversity of choices, especially when dealing with international markets for both crude and refined products.

“The question is not to export one or the other,” Lippe stated. “Exports will be anything. Whatever makes the most money, whatever the market wants. The energy companies and traders need to think of new ways to manage their businesses. The approach should not be wide open: Come and get it!”

In some cases, Lippe said, the majors are likely to be exporting to themselves.

“For example, people talk about how much oil Singapore imports. Singapore itself hardly uses anything. It is the refineries around Singapore that import the crude. And who owns those refineries? Some of the same majors that will be exporting from the U.S.,” he said.

In effect, what Lippe is suggesting is re-intermediation. After half a century of disintermediation—when the integrated companies sold off their retail and distribution networks, even refineries and tanker fleets—it now may make sense to coordinate those segments to some degree again.

Private equity
The wildcard in plans for new export terminals is private-equity money.

“There is so much of that flowing around looking for projects that all reasonable proposals can be funded. Look at the names that have been circulating: Enterprise, Marathon, Magellan. Those are all reasonable companies. Any terminal plans by names like that have got to be taken seriously,” he added.

One way or another, Lippe said, the U.S. has to add about 5 MMbbl/d of export capacity for both crude and refined products in the next seven years, “which is feasible. Saudi Arabia exports 7 MMbbl/d to 8 MMbbl/d from just two ports. China imports about 8 MMbbl/d, so the scale that is needed for the U.S. is not a big deal.”

In the meantime, major trunk lines in North America are moving ahead. Lippe expects Keystone XL to be in service in 2019, adding 800 Mbbl/d of transportation capacity to the Gulf Coast. He also expects the lines from Alberta to the coast of British Columbia to go ahead, despite the latter province’s resistance.

“Most of the First Nations through whose land the lines will run are behind the projects, and under the Canadian constitution First Nation sovereignty is strong. That has been litigated over and over through the years, and the tribes have 140 wins and no losses. The British Columbia problem will be solved, and Alberta can fulfill its potential,” he said.

Flexibility and diversity

It is impossible to gauge which projects will eventually fly, Hedge said.

“It will hinge on the economics and who can hit their subscription marks fastest. However, the Oiltanking/Enbridge/Kinder Morgan terminal provides the most flexibility/diversity of supply. It is reasonably close to shore and doesn’t require a lot of on-shore pipe development,” he said.

Jupiter appears to be all but a done deal, Hedge added, “but is a significant development. And while Trafigura has good access to well-positioned assets, it is facing opposition from the Port of Corpus Christi. Tallgrass will have to compete with LOOP. At this point, my money is on Oiltanking/Enbridge/KM and Jupiter, Tallgrass and Trafigura are less likely, and with Enterprise potentially coming in from behind.”

The plan by Oiltanking, Enbridge, and Kinder Morgan is to build a new 10 MMbbl crude storage terminal in Freeport, Texas, connecting to a new offshore loading facility by 2022. The loading terminal would be about 30 miles offshore. The terminal would be connected with the Gray Oak Pipeline from the Permian to Corpus Christi. That is a 700 Mbbl/d line planned by Phillips 66 Partners and Andeavor, which recently combined with Marathon Petroleum. Kinder Morgan has an existing crude pipeline from the Eagle Ford to Sweeny, Texas, near Freeport and potentially to the Seaway system from Cushing in which Enbridge holds a 50% interest.

The Jupiter Offshore Loading Terminal (Jolt) is an add-on to a greenfield project to build out a crude/condensate upgrading facility, fuel blending facility, and an import/export marine terminal for crude and refined products. It will be supplied from a proposed 670-mile, dedicated high-gravity crude pipeline that originates in Orla, Texas, with additional injection and offtake points at Pecos and Three Rivers, Texas.

“If you assume that the pipeline and onshore terminal is a done deal,” Hedge said, “then adding about a five-mile pipe to deepwater and the loading platform is a minor addition. The advantage to being in Brownsville is that deepwater is only 5 miles offshore, as opposed to the Oiltanking/Enbridge/KM proposal which is about 30 miles offshore or Tallgrass which is more like 80 miles offshore. At about $10 million per mile to develop, the cost savings is significant.”

The Tallgrass/Drexel-Hamilton proposal “is much more than a pipeline,” Hedge added. “It includes plans for a new onshore terminal called the Plaquemines Liquids Terminal (PLT). The Seahorse Pipeline will have the capacity to transport 800 Mbbl/d. The PLT will load post-Panamax vessels (up to 1 MMbbl) and be operational by the second quarter of 2020. It will also incorporate a separate offshore pipeline extension that would give PLT the added capability of loading VLCCs by third-quarter 2021.”

Farther in the distance

Trafigura’s Texas Gulf offshore crude export terminal would be about 15 miles off the coast of Corpus Christi and includes an onshore storage terminal connected to an offshore loading terminal by two parallel, 30-inch-diameter pipelines. It would have capacity of 500 Mbbl/d to load VLCCs. The terminal will be linked to Plains All American’s planned Cactus II crude pipeline. Plains subscribed to about 300 Mbbl/d. Plains holds 150 Mbbl/d on the existing Cactus line.

“This project is facing opposition from the Port of Corpus Christi,” Hedge noted. “The port is looking into dredging and widening its ship channel to accommodate VLCCs in the Harbor Island area.”

Enterprise is “aggressive at expanding their footprint are doing their FEED work,” Hedge said. “They have not announced a location although speculation and their existing footprint would suggest Houston. If they go to an open season, then it will become much more realistic.”

Magellan is focused on expanding pipeline capacity, he noted. “They are, through their Seabrook Logistics venture with LBC Tank Terminals, expanding their crude/condensate storage capacity and adding a Suezmax dock to the Seabrook terminal. It would be logical for them to be considering it, but I haven’t heard that they are seriously looking into it.”

In a final note, Hedge advised to “follow the money. If there is a need for capacity then capacity will generally be built, so I don’t believe that any corridors are being neglected from a takeaway capacity standpoint. I think it is important to keep an eye on the natural gas and NGL processing and takeaway capacity—because lack of that can curtail crude production.”

What jumps to his mind are EPIC Midstream’s planned 590 Mbbl/d crude line from the Permian to Corpus Christi; and Energy Transfer Partners’ 600 Mbbl/d crude line from Midland to Nederland, Texas. There are also additions planned to several expansions to existing systems.

Gregory DL Morris is a freelance writer based in Chapel Hill, N.C., specializing in energy and petrochemical topics.