More dangerous than coronavirus? The safety case for digital and remote operations.

Remote working, digital de-manning, drones and robotics — all of these themes will structurally accelerate in the aftermath of the COVID crisis. Our research outlines their economics and how they can accelerate the energy transition. But this short note considers the safety consequences. They are as significant as COVID itself. And equally worthy of re-casting behaviours, policies and investments.


At the time of writing, the United States has been hardest hit by the COVID crisis out of any country in the world. It has incurred c35,000 fatalities. However, in the past five years, the US has also incurred an average of 35,000 fatalities on its roads each year (below). This is c100 deaths per day. 1 out of every 10,000 people is killed on US roads each year. There are 1.2 death for every 100M vehicle miles driven (and 3.2 trn miles are driven each year).

Likewise, at the time of writing, the US has been hit by 700,000 COVID cases. For comparison, there are 2.6M injuries on US roads each year, and 6.3M traffic accidents. This means 1 out of every 125 people is injured on US roads each year. There are 83 injuries for every 100M vehicle miles driven.

If you believe in working from home to save lives amidst the coronavirus crisis, a similar argument may justify working from home, where possible.

In addition, 5,250 US workers were killed in workplace fatalities in the most recent annual data, equivalent to 1 out of every c30,000 full-time employees. 40% of these deaths occur on roads. Of all the major job categories shown below, the most dangerous is trucking, where 1 out of every 4,000 full-time employees is killed each year.

Looking more granularly, COVID has so far killed 1 out of every 10,000 people in the United States. However, fatality rates range from 1 in 10,000 to 1 in 1,000 for workers in some of the more physically intensive industries (as shown below), which comprise 10% of all the hours worked around the US economy.

Workplace injury rates are 3% across the entire US economy. This is also 10x higher than the number of documented COVID cases so far in the United States.

If you believe in using technology to save lives amidst the coronavirus crisis, a similar argument may justifying greater deployment of autonomous technologies, digital de-manning, drones and droids, across the broader US labor market.

Our research finds that 48% of recent digitization initiatives have materially improved safety (chart below). 60% also materially lowered costs, 55% materially increased output and 24% materially lowered CO2 emissions.

To recycle an example from the note, there is no need for a worker to be placed into harm’s way — climbing a scaffold to inspect a roof or lowered on a harness to inspect the undersides of an oil platform — as remote monitoring, drone and robotics technologies become available. This is why we have recently screened which operators are among the technology leaders, including in digital technologies (chart below).

The importance of remote work, digitization technologies and robotics may sound obvious when framed in the terms above. But they are not being deployed sufficiently. The chart below shows the number of road fatalities in the US, declining at a 3.4% CAGR since 1920. But there has been no progress in the past ten years since 2009. The absolute count of road fatalities in the latest data is no better than in 1960 (below).

Likewise, workplace fatality rates deflated at 3% pa since 1992, but they have also since stalled. No net improvement has occurred since 2009.

Safety matters, during the COVID-crisis, and after the COVID crisis. Remote and digital technologies can play an enormous role, if enabled by policies and embraced by forward-thinking companies. Please contact us if we can help you screen opportunities. And sorry for the morbid tone of this short note.

The new forest: can carbon-neutral fuels re-shape the oil industry?

Integrated oils have a game-changing opportunity in seeding new forests. They could potentially offset c15bn tons of CO2 per annum, enough to permit the continuation of 85Mbpd of oil and 400TCF annual gas consumption within a fully decarbonized energy system. The cost is competitive, at c$50/ton. It is natural to sell carbon credits alongside retailing fossil fuels. We calculate 15-25% uplifts in the value of a typical fuel retail business, while allaying fears over the energy transition. Our 21-page note outlines the opportunity.


The advatages of forestry projects are articulated on pages 2-5, explaining why fuel-retailers may be best placed to commercialise genuine carbon credits.

Current costs of carbon credits are assessed on pages 6-8, adjusting for the drawback that some of these carbon credits are not “real” CO2-offsets.

The economics of future forest projects to capture CO2 are laid out on 9-10. We find c10% unlevered IRRs at $50/ton CO2 costs.

What model should fuel-retailers use, to collect CO2 credits at the point of fuel-sale? We lay out three options on pages 11-14. Two uplift NPVs 15-25%. One could double or treble valuations, but requires more risk, and trust.

The ultimate scalability of forest projects is assessed on pages 15-19, calculating the total acreage, total CO2 absorption and total fossil fuels that can thus be preserved in the mix. Next-generation bioscience technologies provide upside.

A summary of different companies forest/retail initiatives so far is outlined on page 21.

Drones & droids: deliver us from e-commerce

Small, autonomous, electric delivery vehicles are emerging. They are game-changers: rapidly delivering online purchases to customers, creating vast new economic possibilities, but also driving the energy transition. Their ascent could eliminate 500MTpa of CO2, 3.5Mboed of fossil fuels and c$3trn pa of consumer spending across the OECD. The mechanism is a re-shaping of urban consumption habits, retail and manufacturing. The opportunities are outlined in our new, 20-page report.


The average US consumer buys 2.5 tons of goods per year, served by a vast distribution network of ships, trucks and smaller vehicles, collectively responsible for 1.5 barrels of oil, $1,000 of cost and 600kg of CO2 per person per annum (page 2).

Fuel economy currently deteriorates, with each step closer to the consumer. Container ships achieve c900 ton-miles per gallon of fuel. But delivery vans, the dominant delivery mechanism for internet purchases, are least efficient, achieving just 0.02 effective ton-mpg and costing at least $3.6 per delivery (page 3).

The rise of e-commerce has already increased supply chain CO2 by c30%, and supply chain costs by 2x since the pre-internet era. On today’s technologies, CO2 will rise another 20% and cost will rise another 50% by 2030, adding 0.7Mbpd of oil demand, 120MTpa of CO2 and $500bn of cost across the OECD (pages 4-5)

Drones and droids are 90-99% less energy intensive than delivery vans, and 70-97% less costly. The technology is maturing. Thus small, autonomous, electric vehicles will move immediately, efficiently, straight to their destination (pages 6-8).

Retail and manufacturing will have be transformed by the time drones approach 50% market share in last-mile delivery. Tipping-point economies-of-scale mean that they will take market share away from cars and delivery vans very rapidly (pages 9-10).

The second half of the report focuses in on the opportunities. Retail businesses must consolidate, specialise or diversify to “sharing” models. The latter can save $1trn of consumer spending and 100MTpa of emissions in the US alone (pages 11-20).

Is gas a competitive truck-fuel?

We have assessed whether gas is a competitive trucking fuel, comparing LNG and CNG head-to-head against diesel, across 35 different metrics (from the environmental to the economic). Total costs per km are still 10-30% higher for natural gas, even based on $3/mcf Henry Hub, which is 5x cheaper than US diesel. The data-file can be downloaded here.

The challenges are logistical. Based on real-world data, we think maintenance costs will be 20-100% higher for gas trucks (below). Gas-fired spark plugs need replacing every 60,000 miles. Re-fuelling LNG trucks requires extra safety equipment.

Specially designed service stations also elevate fuel-retail costs by $6-10/mcf. Particularly for LNG, a service station effectively ends up being a €1M regasification plant (or around $250/tpa, costs below).

We remain constructive on the ascent of gas (below), but road vehicles may not be the best option.

To flex our input assumptions, please download our data-model, comparing LNG, CNG and other trucking fuels across 35 different metrics .