Its been 10 days since Keystone Pipeline was shut down and its hard to see a quick restart. We have a regulator with responsibility (Pipeline & Hazardous Materials Safety Administration) for the pipeline restart and one that wants to make sure it has its say (South Dakota Public Utilities Commission). There is no formal indication for the cause of the leak, but it sounds like it is different than the cause of the nearby April 2016 Keystone Pipeline leak. The only real option for displaced volumes is by rail as Enbridge’s heavy oil lines are full. But we believe Gulf of Mexico refineries will want to lock up Cdn heavy oil barrels because of declining Venezuela and Mexico oil production. Just like Venezuela and Mexico led to narrower than expected Cdn heavy oil differentials in 2017, it should help keep a cap on the widening Cdn heavy oil differentials to not much more than the cost of rail.
Its taking longer to restart Keystone than most originally expected. Its now been 10 days since TransCanada shut down the Keystone pipeline in the early morning hours of Nov 16. Many of the original expectations were for a quick fix and restart wit a restart expected in 1 to 2 weeks. TransCanada’s latest posted update (530pm CST Nov 24[LINK]) notes “Preliminary work to expose and subsequently extract the damaged section of pipe began yesterday and is expected to be complete by end of day Sunday, November 26. Additional excavation will be conducted beyond Sunday for soil remediation purposes. Preliminary inspections of the damaged section will be completed on site by both TransCanada and PHMSA staff, then sent to Washington, D.C., for a complete investigation by the National Transportation Safety Board’s Metallurgical Laboratory.” Its positive that it sounds like they are getting close to finding out what caused the leak. And the cause of the leak and determining if this causes other risks to the pipeline will be the key factor in determining a restart. The timing concern is that they are sending the preliminary inspections for a “complete investigation” in the metallurgical laboratory. PHMSA can move quickly as seen in the April 2, 2016 Keystone Pipeline link near Freeman (southern part of South Dakota), when PHMS issued its corrective action order within 7 days of the leak, on April 9, 2016 [LINK]. The timing concern is this oil leak is that if feels like sending the preliminary inspections for a “complete investigation” in the metallurgical laboratory isn’t a turnaround in a couple of days.
The key regulator for the restart is the PHMSA, not the South Dakota PUC. The TransCanada update reminds that the key regulator for the restart is the Pipeline & Hazardous Materials Safety Administration (PHMSA) from the US Department of Transportation. The PHMSA makes the decision to restart following its review of the cause of the leak, the fixing of the leak and the ability of the pipeline to operate in a safe manner going forward. It probably doesn’t surprise anyone that the PHMSA has stayed out of the media spotlight on the Keystone oil leak. We think any of the pipeline companies view it as a positive that PHMSA has the regulatory responsibility
But the South Dakota Public Utilities Commission (PUC) is the likely the bigger wildcard even if PHMSA signs off a restart. Our Nov 22, 2017 blog “Tough Regulator Comments Make It Seem Like The Keystone Oil Pipeline Shut Down Isn’t Days, Or A Couple Weeks, But Much Longer” [LINK] noted the quoted comments from all three PUC commissioners that clearly suggest the PUC will be looking at this incident with a tough eye. The PUC is also being careful to note its concerns are linked to its regulatory authority on the original permit to construct in South Dakota ie. “We are waiting to see what the forensic analysis comes back with to see if any of our conditions were violated,” and “The PUC needs to determine whether any of the permit conditions for this pipeline were violated. Those conditions were placed on the permit to ensure safe construction and operation of the pipeline”. We can’t assume PHMSA will be a rubber stamp to approve a restart, but it seems like the PHMSA is much more technical review process. Whereas we worry that that the PUC is the bigger wildcard, they are clearly saying they want to ensure there has been no violation of the original permit ie. a separate process from the PHMSA. The PUC sensitivity is likely because of the Keystone April 2016 oil leak that was south of this year’s leak. And we also worry what happens if PHMSA gives the okay to TransCanada to restart and the PUC says hold on.
The other reason why we worry about the length of time is that it doesn’t sound to be the same cause as the April 2016 oil leak. As noted earlier, PHMSA moved quickly to corrective action in April 2016. At that time, PHMSA said “On April 7, the immediate cause of the leak was identified as a girth weld anomaly at the 6:00 position on a transition weld. The cause of the welding anomaly is still under investigation and unknown at this time. A third-party metallurgist, contracted by TC , discovered the weld anomaly. At the time of the discovery, the anomaly was observed to be releasing crude oil from the pipeline at a rate of approximately two drops per minute. It is not known how long the pipeline has been leaking”. We were a little surprised at some reported TransCanada comments to the local media. On Nov 21, ABC affiliate in Sioux Falls (KSFY) reported [LINK] “TransCanada: Keystone oil leak a sudden, ‘immediate’ event” and “A TransCanada Corp. official says he believes an estimated 210,000-gallon oil leak discovered last week in South Dakota from the Keystone pipeline was a sudden and “immediate” event. The American News reports that Erik Tatarchuk, a TransCanada vice president, said at a Marshall County Commission meeting Tuesday that it is unlikely oil leaked long enough to soak into the soil.” The description of the leak being sudden and immediate makes it sound something abrupt and not like the slow leak in Apr 2016.
There is no near term heavy oil by pipeline solution until Enbridge Line 3 expansion in H2/19. The Cdn heavy oil producers don’t have any pipeline options, at least until the Enbridge Line 3 expansion’s planned start in H2/19. It is planned to add 375,000 b/d of added heavy oil capacity and is what we have always considered as the most likely soonest relief valve for Cdn heavy oil.
But a pickup in heavy oil by rail should happen with increasing Mayan less WCS differentials. Enridge is already running full so can’t take more Cdn heavy oil. Enbridge’s apportionment on Cdn heavy oil in Dec is reportedly around 20% ie. there were nominations for heavy oil capacity 20% greater than the capacity. But crude by rail is now an economic option to the Gulf of Mexico. There will be logistical issues like getting the tank cars and locomotives in place. The good news is that Cdn heavy oil industry had geared up previously for crude by rail so there is a lot of loading/terminal infrastructure already in place. The Keystone Pipeline shut down has led to an immediate widening of the Mayan less WCS differential. This is critical as it now sets the stage for more rail. Most generally use a $14 or $15/b cost of rail for WCS to reach the Gulf of Mexico refineries via rail. The Mayan less WTI differentials widened to $14.43 on Thursday, before pulling back slightly to close at $13.90/b on Friday – basically at or close to the breakeven price for refineries in the Gulf of Mexico. This sets up the potential for rail to jump in to get Cdn heavy oil to the Gulf of Mexico.
Mayan Less WCS Differential US$/b
Source: Bloomberg
The key for Cdn heavy oil differentials remains the declining Mexico and Venezuela oil production. One of the key reasons for narrower than expected Cdn heavy oil differentials in 2017 has been declining Mexico and Venezuela heavy oil production. These narrow differentials have been the big plus in Cdn oil and gas industry. Historically, Mexico and Venezuela have been the primary suppliers of heavy oil to the Gulf of Mexico refineries. The Mexico and Venezuela production declines in 2016 have not changed in 2017. This is why we believe Gulf of Mexico refineries will be doing what they can to lock up Cdn heavy oil supply and this will help keep a cap on differentials not too far from the cost of rail.
Mexico Oil Production (thousand b/d)
Source: Pemex
Venezuela Oil Production (thousand b/d)
Source: OPEC
The Fort Hills startup is the wildcard that can widen the differentials. Suncor’s For Hills project is scheduled to commence operations in Dec and ramp up to “achieve 90% of its planned production capacity of 194,000 barrels per day within 12 months”. This will add another ~180,000 b/d of heavy oil in a market that is currently short of pipeline capacity. If the Keystone Pipeline is shut down for months instead of weeks, we would expect to see the additional heavy oil volumes from Fort Hills likely force a back out of some heavy oil volumes and lead to a widening in the Cdn heavy oil differential. One of the big mysteries in Cdn heavy oil capacity is how different is the actual capacity vs nominated volumes to pipelines. We believe the actual capacity is less in time like today, we just don’t know how much less. One other winter factor will be weather as cold weather can impact rail operations.
It’s hard to see a quick restart to Keystone Pipeline, heavy oil by rail is the only near term alternative. It’s been 10 days since Keystone was shut in. PHMSA has shown they can respond quickly but we think this is still pointing a longer than expected shut in given it doesn’t look like a slow small leak as caused the April 2016 Keystone pipeline leak. We may not believe the PUC revoking of the permit is a likely event, however the PUC seems determined to play a key role in the Keystone restart and this is the regulatory wildcard. Enbridge’s heavy capacity is full and their Line 3 expansion is ~18 to 21 months away. The widening Mayan less WCS differential is basically at or near the cost to move Cdn heavy oil by rail to the Gulf of Mexico so look for heavy by rail to pick up. Most of all, we believe Gulf of Mexico refineries will want to lock up Cdn heavy oil given the continuing decline in Mexico and Venezuela heavy oil production. This is the key reason why Cdn heavy oil differentials shouldn’t be much more than the cost of rail. At least until we see how the Fort Hills start up impacts the equation.