We believe there is the risk for Cdn heavy oil differentials to be hit harder than the $1.55/b so far since the Keystone pipeline was shut down on Nov 16. As of 4pm MDT, TransCanada (TRP) has not disclosed the findings for the cause of the Keystone pipeline leak, the expected time to clean up the spill, repair the pipeline and then restart the pipeline. TRP has not inferred any bigger problem or bigger issue. Primarily based on the straightforward TRP limited updates and aerial pictures, the market is believing the cleanup and pipeline repair can be accomplished in a matter of days not weeks. We share that view. However, we think a pipeline restart is likely to take longer than expected. Its not just the clean up and repair for regulators to sign off, regulators will need to be confident in the cause of the leak and that it isn’t likely to happen elsewhere. We think the regulator is likely to take more time on this restart, especially given the Keystone name and last year’s Standing Rock protests that delayed the Dakota Access Pipeline. Our worries are that a restart that takes 2 weeks or more will lead to Cdn heavy oil diffs being hit harder than the $1.55/b to date.
Cdn heavy oil differentials were immediately hit by the Keystone pipeline leak. Cdn heavy oil prices reacted immediately from TransCanada’s announcement that Keystone pipeline was being shut down for the spill following the oil spill. This immediately led to a shut in of Cdn heavy oil, and it also led to an immediate increase in the discount for Cdn heavy oil. The WCS less WTI differential increased from $14,20/b on Thus to $15.75/b today. Keystone has a capacity of 590,000 b/d.
WCS Less WTI Differential US$/b
Source: Bloomberg
TRP has not yet determined the cause of the pipeline leak. TRP set up an Amherst Incident site for the oil spill and TRP’s response [LINK]. The key determinant to how long the pipeline will be shut down will be for TRP to figure out what caused the leak. Without knowing the cause, TRP can’t put forward a position to the regulators for future assurances. On Sunday, TRP advised “We are continuing to make progress in our investigation to determine the cause of the leak on the Keystone Pipeline right-of-way”. As of our signoff on this blog at 4pm MDT, TRP has not posted any update on the findings for the cause of the leak.
It looks like TRP’s emergency shutdown procedures worked well. In its Nov 16 update, TRP noted “At approximately 6 a.m. CST (5 a.m. MST) today, we safely shut down the Keystone pipeline after we detected a pressure drop in our operating system resulting from an oil leak that is under investigation. The estimated volume of the leak is approximately 5,000 barrels. The section of pipe along a right-of-way approximately 35 miles (56 kilometres) south of the Ludden pump station in Marshall County, South Dakota was completely isolated within 15 minutes and emergency response procedures were activated.” Every pipeline has an operating control room that allows it to shut down the system – basically like a stop the assembly line button, it’s a way to shut down the system in whole or in part at certain designated points. It seems that TRP’s emergency shutdown worked if there were only 5,000 barrels leaked, which infers the system was shut down very quickly after seeing the pressure drop in the pipeline. We were reminded by our key pipeline contacts today that the size of the leak depends upon the time to shut down and how far apart the sectionalizing valves are and therefore how much crude may continue to seep out of the ruptured area once the line is shut down. The pressure drop indicators will identify the pipeline leak location, the operator will immediately shut down the valve ahead of that lead. The valve shutdown means that no more oil flows into the problem part of the pipeline. Not all of the oil in the pipeline will necessarily leak out, but the oil seeping out is therefore a function of how much more oil seeps out after the valve is closed. At 5,000 barrels leak, that theoretically represents 12 minutes of oil based on 590,000 b/d daily capacity. But the valve shut down likely happened with a minute or two. If you look at a calculation of barrels of oil per foot of pipeline, it would show that for a 30” pipeline like Keystone, a pipeline calculation would show that there is about 1 barrel per 0.83 feet of pipeline. That sounds reasonable, this means that every mile of pipeline would hold 4,400 barrels of oil vs the reported 5,000 barrels leaked. Assuming that not every barrel of oil per foot leaks out, the implication is that the nearest valve aren’t likely too far away and TRP did a quick shut down ie. their emergency shut down procedures worked.
On the surface, it doesn’t seem like it should take very long for TRP to clean up and fix the pipeline. We can see why the market believes this leak can be dealt with quickly. TRP’s updates suggest a straight forward leak ie. nothing ominous. TRP’s update said “we currently have all the necessary equipment for the remediation work, including dump trucks, excavators and bulldozers. Measures continue to be in place to deter wildlife from entering the area and sophisticated air-monitoring equipment is being used around-the-clock – though there have been no significant concerns.” And based on the limited aerial pictures, we would share that view that a cleanup and pipeline repaid should not take too long. Unfortunately, TRP has not given any detailed explanation of the spill, the clean up procedures or the time to clean up and fix the pipeline. Its also important to note that there are not hints or inferences by TRP that this is a major oil spill or that there is something more significant that what is being observed. As a result, the market has looked at aerial to try to rationalize the extent of the leak and is coming up with a view that not likely a big issue. Reuters posted Nov 16 aerial pictures of the spill area [LINK] . The pictures don’t show any significant pools of oil on the surface or that are looking to flow into the ditch to the right side of the picture. Plus, the reports today are that no oil seeped into the ditch on the right side of the picture due to a berm. Rather the pictures seem to portray that the oil has leaked out of the pipeline and soaking the soil. On the surface, this would suggest a reasonably straightforward clean up – dig it up. And on the surface, it shouldn’t be a logistical problem to a quick clean up in a matter of days not weeks. Plus repairing a section of a pipeline is also a quick process, something that is done in a matter of days.
Keystone Pipeline Leak – Nov 18, 2017 Aerial Pictures
Source: Reuters
But we think the market may be overlooking a big wildcard – TRP will need regulatory approval to restart. We are more cautious on a quick restart and believer the big question mark on time to restart will be the regulator. It shouldn’t take long for a regulator to sign off that TRP has cleaned up the oil leak properly and fixed the pipeline leak. However, the regulator needs more to allow a restart. The regulator will need to have confidence in the reason for the pipeline leak is correct, and that the reason for the leak is unlikely to happen again somewhere else. We are less worried about delays from US Thanksgiving holidays but our worries that a restart of an pipeline named Keystone in this area has heightened sensitivity for regulators. The name Keystone will be staying in the news with the neighbouring Nebraska Keystone XL approval today that are only starting the first round of appeals. These groups are inevitably going to focus on the Keystone spill. Plus the famous Standing Rock protest site that delayed the Dakota Access Pipeline for oil spill fears was just across the border approx. 150 km away at Cannon Ball North Dakota. It just seems that, even under a straightforward cause and explanation, the regulators aren’t likely to make a quick restart decision. We hope we are wrong, but we believe it will be tough for a quick restart. And it is easy for a couple weeks to pass.
If a restart drags, we see the risk to Cdn heavy diffs being hurt even further. No one but TRP has the inside information on the findings to date. Based on the limited observations and no inference from TRP that there is a bigger deeper problem, we are in the camp with everyone else that a clean up and fixing of the pipeline isn’t likely to be a long process. However, we are different in that there is a big wildcard to get regulatory approval for a quick restart even if it is a simple problem, clean up and fix. Our worries are that a restart that takes 2 weeks or more will lead to Cdn heavy oil diffs being hit much harder than the $1.55/b to date.