Letter of Comment October 2016

Who we are

Concerned Professional Engineers (CPE) is a group of Registered Professional Engineers. We have extensive experience in the design, operation and maintenance of resource export terminals, design of escort tugs, handling of ships and navigation.  We are not unconditionally opposed to the shipment of resources overseas, as we believe that we have an ample supply in Canada.  We also believe that we have generally been responsible at exporting resources in an environmentally sensitive manner and that export of resources is a vital part of our economy. Nevertheless, Canada should emphasize the export of value-added products rather than just the shipping of raw resources.

We have examined the marine aspects of the Kinder Morgan Trans Mountain Expansion (TMX) project and found that the planned increase to transport Diluted Bitumen (Dilbit) from Kinder Morgan’s existing terminal in Burnaby, through the straits of Georgia and Juan de Fuca to the Pacific Ocean, presents a potential high risk to the environment and built infrastructure.

Project Risk

Our main purpose in writing this Letter of Comment on the TMX is to point out that the risks of an accident and a bad spill from the increased traffic are considerable.  Based on Trans Mountain’s own experts’ estimations, as submitted with their TERMPOL Report 3.15 (Table 34), of November 25, 2013, there is a ten percent (10%) probability that a spill of 8.25 million liters or more will occur in a 50 year operating period, even with all the proposed mitigation strategies. This is considerably greater than the mitigated spill risk of nine percent for a 5.0 million liters spill estimated for the Northern Gateway project out of Kitimat. The probability of at least one spill in 50 years increases to 19% when spills of any size are considered.  CPE  does not have access to the model that KM’s experts (DNV) used to predict the probability of spills. This model should be made available and a completely independent analysis of the spill risk should be carried out.

We also believe that there has not been a proper analysis of the potential for a collision of fully loaded or empty Aframax-type tankers with either the First or the Second Narrows bridges, particularly with the present Second Narrows railway bridge. Aframax is a medium-sized crude tanker with a dead weight tonnage (DWT) ranging between 80,000 and 120,000 Tonnes, and a length of 245 meters. The regional economic consequences of bridge damage (or collapse) following a collision accident cannot be over-emphasized. The history of collisions of vessels with these bridges needs to be carefully analyzed and re-evaluated with regard to the proposed TMX traffic.  This review should include the number of times the railway bridge has been knocked-out of service for a considerable amount of time and the number of times that it has had to be completely replaced.  We believe that when this analysis is done, the risks will probably be considerably higher than those stated by Kinder Morgan’s experts.

In our opinion, there needs to be an analysis as to what would happen if there is a collision of a loaded Aframax vessel with the Second Narrows railway bridge or highway bridge.  What forces would be exerted on the bridge structure or foundations and what would be the expected damage to these bridges?  Also, would the forces  exerted by the vessel in striking the foundations of the bridge be sufficient to damage or ripping a double-hulled vessel, resulting in a release of its oil cargo?

We want to refer you to a paper by Dr. Ricardo Foschi, P.Eng and Emeritus Professor of Civil Engineering at the University of British Columbia. He considered these vessel collision probabilities and prepared a set of questions that need to be answered by the project proponent. This analysis, based on the requirements by the Canadian Highway Bridge Design Code (CAN/CSA S6), is attached as an Appendix to this letter. The analysis shows that the probability of collision with the bridges is very much dependent on the effectiveness of tugboat assistance in case a tanker is out of control. We think that a detailed modelling of the interaction dynamics of the system tanker-tugboats must be shown, and that this model should be used to estimate the degree of control that the tugboats can achieve. We are aware, for example, that similar studies have been carried out by the city of Seattle for the tanker traffic in Puget Sound, or by San Francisco for similar traffic in Northern San Francisco Bay. We find that the proponent has not offered a similar detailed modelling of the effectiveness of tugboat aid in relation to the existing infrastructure along Burrard Inlet.

We believe that increasing Kinder Morgan (KM) tanker traffic through the heart of the Port of Vancouver should be seriously evaluated vis-à-vis other alternatives. For example,  Roberts Bank Superport was  built for the purpose for handling large cargo ships and we believe that it would be a much safer alternative for shipping Trans Mountain’s export Dilbit.  Roberts Bank is safer because it is much closer to the open ocean, and does not have the obstacle course presented by the First and Second Narrows bridges. The question needs to be asked: why is the product not proposed to be shipped through Roberts Bank?  This question needs to asked of the Port of Vancouver as well. It also seems that restricting vessel size to Aframax class is unnecessarily constraining.  At Roberts Bank, VLCC-class tankers could be used that offer three times the capacity or requiring  one third the number of ships. The pipeline transportation corridor to Roberts Bank is also available along the Roberts Bank Coal Traffic rail right-of-way.

The product being shipped 

The behaviour of Dilbit in seawater as has been the subject of much debate. There is no clear evidence that should a spill occur, and depending on the sea conditions the product will stay on the surface long enough for it to be cleaned up. It can be safely said, however, that cleanup costs of a Dilbit spill will be very large.  The cleanup and compensation cost of seven billion dollars, attributed to the Exxon Valdez Alaska  incident 25 years ago, may be a low approximation to the requirements for a spill in the Kinder Morgan project. It is even possible to speculate that Kinder Morgan may want the oil that is being spilled to sink, so that it is out of sight and out of mind.

The project proponent must be asked to produce scientific evidence on the behaviour of Dilbit under all sea conditions and on these basis, produce a realistic clean-up response strategy. The company needs also to produce a scientific assessment of any spill consequence related to the toxicity of the product.

Liability for cost

It is clear that of the owner of the tanker, not Kinder Morgan, is liable for a spill cleanup and compensation costs. We believe that the funds available according to the latest estimates of the Federal Government are $1.3 billion, which would fall way short of cleaning up and compensation for a 8.5 million liter (or greater) spill. It is likely that the Dilbit will separate from the condensate that enables it to float,  and that it will  then sink and form  tar balls which, over the years, will make its way to Greater Vancouver’s    shores, triggering a continuing clean up and compensation  mess. It is our view that the product should be upgraded in Alberta and then shipped as a light crude.

The liability fund estimated by the Federal Government is a long way from covering the actual costs for cleanup and compensation. In our view Kinder Morgan should require all the vessels that come to pick up the product to have unlimited liability insurance. If this were the case, the insurance company would do a realistic assessment of the risks and would increase the premiums. Therefore these premiums would be added to the cost of the barrel of oil and we would see a more realistic cost of the price of oil.

We urge you to consider these matters very carefully, and thank you for allowing us the opportunity to submit a Letter of Comment on this very important project.

Yours sincerely,

Brian Gunn

Spokesperson for Concerned Professional Engineers.









Ricardo O. Foschi, P.Eng

October 2014


Kinder Morgan presents calculated return periods (in years) for oil spills of different volumes. These are given in Table 34 of their TERMPOL Report 3.15. The oil spills result from marine accidents or incidents.

By definition, the return period is an estimation of the average time elapsed between spills of a given volume. As such, approximately 50% of the spills would occur before the return period and 50% would occur after that time. Therefore, the return period is not a good statistic to communicate probability of a spill. Of importance is the probability that at least one spill, of a given volume, would occur within the operating life of the project.  The calculation of such a probability is straightforward given the return period and the operating life.

Table 1 below shows the results of this calculation, starting from the Kinder Morgan estimations.

Oil spill volume (m3) Return Periods (in years) Probability of at least one spill in 50 years

No projectProject with no mitigationsProject with all mitigations

No projectProject with no mitigationsProject with all mitigations>16,5003,0934562,3660.0160.1000.020>8,250619914730.0800.4200.100>0.0 (any)309462370.1500.6600.190


The above Table permits the following conclusions:

  • With no mitigations the probabilities of at least one oil spill in 50 years are too high. Thus, mitigations are essential and must be enforced.
  • Even with mitigations, probability of at least one oil spill in 50 years, greater than 8,250 cubic meters, is deemed to be too high (0.10 or 10%). This is comparable to the probability for a spill greater than 5,000 m3 calculated for Northern Gateway (9%). The probability for a large spill of 16,500 m3 is more tolerable (0.02 or 2%), but even a more moderate spill would cause very substantial damage.
  • Even with mitigations, there is a 19% probability of an oil spill, regardless of volume. This is also too high.

The methodologies for the determination of the probability of collision of a vessel with a bridge pier are specified both in the American AASHTO Code (1991) as well as in the Canadian CSA-S6-00 (2000). Both Codes essentially contain the same provisions, differing in the system of units used in the prescribed equations. The methodology followed here agree with that specified in the Canadian Code CSA-S6-00.

This methodology has been used to evaluate the risk of collisions with several new bridges across the Fraser River:  Golden Ears, Pitt River, Port Mann and the Skytrain Canada Line.

The methodology is based on the estimation of:

                  PA = probability of aberrancy, or the probability that a vessel will be out of control

or likely to be involved in a collision incident;

                  PG =  conditional, geometric probability that a vessel will collide with a pier, given that the

vessel is out of control or likely to be involved in a collision.

The product   PE = PAPG  gives the probability that a collision will take place, which has to be modified according to the number N of vessels transiting per year or in any interval T. From historical accident data in US waterways, the Code gives the value  PA= 0.6 x 10-4, applicable to ships.

The geometric probability is calculated considering that the position of the ship in distress is randomly located, with a mean equal to 0.0 (the centerline of the navigation channel) and a standard deviation equal to the length  Ls of the vessel. This random position s is assumed to obey a Normal distribution.

If mitigation aids from tugs were present, then the standard deviation of the position s would be smaller. For perfect mitigation, the tugs would keep the vessel along the centerline of the channel.  In the calculations shown here it is assumed that the standard deviation could be a value  (Ls / r) , with r being a factor either 1, 2, 4 or 6. Thus, r = 6 would imply a more effective mitigation by the intervention of tugs.

The probability PE is finally corrected for the number of vessels transiting the bridge location per year (here assumed to be 600), and then for the period of operations T = 50 years.

The vessel considered is an Aframax tanker, with a length of 245m and a beam of 34m. The opening of the central span of the highway Second Narrows bridge is 350m.

Results are show in the following Table 2:


Factor r PG PE PAnnual (600 vessels/y) Pat least one collision in 50y
r = 1     (no tugs) 0.304 1.9152 x 10-5 0.0114 0.437
r = 2 0.155 9.7650 x 10-6 0.0058 0.254
r = 4 0.027 1.7010 x 10-6 0.0010 0.050
r= 6 0.004 2.5200 x 10-7 0.0002 0.008


It can be concluded, from these results, that the probability of at least one collision with the bridge, over 50 years of operation and at 600 vessels per year, must be mitigated by the use of tugs. This is essential and must be enforced. With proper and effective mitigation, it would appear that collision with the bridge could have a low probability of occurrence.

Collision with the bridge does not necessarily mean major damage or collapse of the structure, nor an oil spill. However, damage to the bridge would result in interruptions of traffic flow with associated economic consequences.  Collapse of the bridge or substantial damage could be studied, but it would require a detailed structural analysis of the bridge and its footings.

  • A more comprehensive model should be studied to relate the factor r to the tug intervention policy.
  • These results apply to the highway Second Narrows or Ironworkers Memorial bridge. The situation for the railroad bridge would be more risky, given that the channel between the bridge towers is much smaller than 350m. For this bridge it would be even more essential to provide an effective mitigation policy.