More on Oil Spills: Some Toxicological Calculations and What if it were Dilbit?

I have been asked a number of questions, both on Twitter and via email, regarding my post on the English Bay fuel spill and so I have written up this follow-up to my last post. In particular, I was asked what the spill would mean from a toxicological standpoint and what would happen if the spill was dilbit rather than bunker fuel.

What Concentration of Fuel Oil will fish be exposed to?

From the toxicological perspective we need to consider a few factors. First and foremost, we all know that any birds or mammals that come in contact with the actual spill itself (get oil covered) will be affected. This is not new so I won’t say any more about that topic. Similarly, any sea life that ingests tar balls or gets covered in oil (like barnacles) will suffer. The question I have been asked most, however, reflects the Shell MSDS for marine oil that has been spread around the internet (ref). The MSDS reports the aquatic toxicity of marine oil as “Very toxic: LL/EL/IL50 < 1 mg/l (to aquatic organisms)” and the question I have been asked is: what will the spill mean to fish and marine life in the area?

The first thing to remember is that MSDSs are not toxicological documents and are intended to provide first responders etc… with very simple information. In this case, the MSDS clearly states “Information given is based on a knowledge of the components and the ecotoxicology of similar products.” If you want to actually know the toxicology of this material it is best to go to Health Canada documents or the US EPA Integrated Risk Information System (IRIS) which the BC Ministry Environment considers the gold standard for risk assessment. The US EPA has a “Robust Summary of Information on Heavy Fuel Oils” (ref) that fills in any gaps you may have about fuel oil toxicology. According to the EPA, fish and crustaceans show effects at concentrations around 1000 mg/L, daphnia (marine plankton) were affected by concentrations of 100 mg/L and algae showed effects (reduced growth rate) down to the 1 mg/L range. With regards to humans, by toxicity standards the material is considered to have low human toxicity with the primary concern being inhalation of vapours containing hydrogen sulphide (from fresh material). That being said, low toxicity doesn’t mean it is good for you, just that it won’t kill you if you accidentally taste the material and thus, I would strongly advise against ingesting the stuff.

Knowing the target numbers to look for, I then did some back-of-the-envelope calculations to establish potential bunker fuel concentrations in English Bay. As described in my previous post, bunker fuel is hydrophobic and almost insoluble and most of the spill has been recovered so all these numbers I’m going to produce are massively conservative (will be much higher than the actual concentrations experienced) but for interest purposes let’s see what the maximum concentrations you could calculate for the spill. I understand as much as 3500 L of fuel escaped into the Outer Harbour. Using Google Earth I established that the Outer Harbour is sufficiently circular in shape that it can be approximated by a circle with an approximate radius of 3500 m. This makes the approximate area of the harbour 3.8 x 107m2. According to the Port of Vancouver, the approximate depth of the anchorage in the Outer Harbour is 14 m. To account for variations in depths, let’s assume that the average depth of the harbour is 10 m. That would make the volume of the harbour approximately 3.8 x 1011 L. As another measure of conservatism, let’s assume that the Outer Harbour represents a big bathtub with no currents bringing water in or out. Assuming all 3500 L mixed evenly in the Outer Harbour the approximate concentration would be 9.2 x 10-9 L/L. This represents a maximum concentration of 9.2 parts per billion. According to the EPA, above, the lowest acute toxicity result for marine toxicity is equivalent to approximately 1 mg/L (1 part per million). Thus, the resulting concentrations from the spill would represent approximately 1 thousand times (three orders of magnitude) lower than the dose that is reported to cause the algae to grow less quickly. [note I had a conversion error in an earlier version of this post which was identified by a sharp-eye reader]. So while the physical exposure to the liquid fuel would definitely have a deleterious effect, initial calculations provide a measure of confidence that a die-off of marine biota would not be expected from this spill since the conservative calculation shows numbers approximately one thousand lower than those that would be expected to effect the most sensitive species in the toxicological tests.

Addendum: a commenter asked what a change to 1 mm thickness would do to the concentration? The difference between 10 m and 1 mm is 4 orders of magnitude, so the resulting concentration would go from 1 million times lower to 100 times lower than those that would be expected to effect the most sensitive species in the toxicological tests.

What if it had been a Dilbit spill?

So the question I have heard the most is: what if were diluted bitumen (dilbit)? I had a lot of thoughts on what would happen having read many older reports on the subject, but rather than trusting my memory I sought out some newer information and after a bit of digging was rewarded when a colleague on twitter (H/T @natrlyst) directed me to an Environment Canada Technical Report on the topic (ref). I have spent several hours digesting its contents (I read this stuff so you don’t have to) and have discovered that much of what we previously believed would happen when dilbit is spilled in the marine environment may be wrong. The following is my précis of that document, with bonus information thrown in. For referencing purposes any physical data/observations should be considered to be derived from the Technical Report, although I will add some additional details.

Let’s start with the basics, what is dilbit? Dilbit consists of a mixture of 20% to 30% diluent and 70% to 80% bitumen. The bitumen is exactly what you think it is and the diluent is typically a naptha-based oil called “condensate”. For those of you who are campers, naptha is essentially the white gas you use in your Coleman lanterns and stoves. The condensate has a specific gravity in the 0.6 g/mL to 0.8 g/mL range and the resultant dilbit has density/specific gravity that ranges from around 0.92 g/mL to about 0.94 g/mL. So the quick answer to your question is that when spilled the dilbit will initially float. We all knew this was the case and up until recently the belief was that as the diluent evaporated away the resulting evaporated mass would sink. Well, apparently, that is not the case. Recent laboratory studies by Environment Canada show that even with a 26.5% evaporation rate (thus with pretty much all the diluent evaporated) the resultant evaporated dilbit would still retain a specific gravity (at 0oC) of 1.021 g/mL. For those of you with long memories you will remember that seawater density ranges from 1.025 g/mL to 1.033 g/mL at that temperature. Thus the material would not sink, as we were previously lead to believe, but would actually remain afloat. More interestingly, typically when lighter oils are hit with breaking waves they form small droplets that lack the buoyancy to float and will often remain entrained in the water column. The dilbit did not act in this way. Rather when the experimental dilbit was exposed to the wave pool, it formed much larger droplets which quickly resurfaced and coalesced into a surface slick. The Achilles heel of the dilbit, however, appears to be sediments in the water. As I mentioned in my previous post, oils exposed to water will preferentially adsorb to sediment particles and will eventually become heavy enough to sink as “tar balls”. When dilbit is exposed to fine to medium sized sediments it does just that and forms tar balls that have a tendency to sink into the water column. Interestingly enough, when the sediment was coarser (sandy grains) the dilbit would not form these tar balls and would instead remain afloat.

So to answer the original question: what if the spill had been a dilbit spill? Well the answer is very interesting because based on the literature, the result would have been very similar to what we saw with the fuel oil spill. Ironically, in the Environment Canada Technical Report they used IFO 180 (a fuel oil mix) as a comparison liquid in many of the experiments and typically it acted in a very similar way to the dilbit. So, believe it or not, the small spill of fuel oil in English Bay provides a very reasonable surrogate for what one would expect from a dilbit spill.

Doing the follow-up research for this piece really reminded me how important it is to keep abreast of the literature on topics such as dilbit spills. While decades of research and experience with oil spills allows us to know with reasonable certainty what will happen in the case of an oil spill, far less is known about the chemistry of a dilbit spill. My initial expectations were shattered as the literature I had on hand (which was only about 5 years old) made a lot of assumptions that have been overtaken by the most recent literature (the Environment Canada Technical Report). Based on this I look forward to seeing what further information is released when the next batch of research studies on dilbit spills hit the literature.

Advertisements
This entry was posted in Chemistry and Toxicology, Oil Sands, Pipelines. Bookmark the permalink.

7 Responses to More on Oil Spills: Some Toxicological Calculations and What if it were Dilbit?

  1. I believe the experiment conditions have to take into account the ambient air temperature, water temperature, wind speed, wave action, and other factors. Did that study you reviewed have variable conditions?

    Like

  2. Pingback: On pipelines, oil-by-rail, and the relative risk of catastrophic spills in the aquatic environment | A Chemist in Langley

  3. CriticalThought says:

    What size are ‘fine’ and ‘medium’ quantitatively? Are they talking clay, and silt or very fine sand, fine sand etc.? I’m sure I can look for myself, but if you have already know this why not just state it for your readers?

    I noticed the report is from 2013, and as you state the importance of keeping up with the literature you may also wish to read the following report: ‘Spills of Diluted Bitumen from Pipelines: A Comparative Study of Environmental Fate, Effects, and Response (2016)’ – http://www.nap.edu/catalog/21834/spills-of-diluted-bitumen-from-pipelines-a-comparative-study-of

    Like

    • Blair says:

      that would be fine to medium marine suspended sediments so would represent clays and silts not sands (which are typically considered in the fine to medium grain range).

      I have a copy of that report. It mostly represents theoretical work and does not do as well as the Environment Canada work in my view.

      Like

      • CriticalThought says:

        >”that would be fine to medium marine suspended sediments so would represent clays and silts”

        Doesn’t all marine water essentially have some fraction of clay particles and silt though? Obviously regions which are more proximal to coastal regions will have a higher fraction of silts and clays than distal areas. Also, pretty rare for calm seas… so we have clay and silts present in a well mixed water column. What’s the average Secchi depth for our coastal water ways?

        Like

  4. Pingback: The Husky oil spill, its effects on the pipeline debate and a thought experiment about a world without fossil fuels | A Chemist in Langley

  5. Pingback: A layman’s guide to the behaviour of diluted bitumen in a marine spill | A Chemist in Langley

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s