
What can satellite imagery tell you about China’s oil policy strategy?
That question was on the mind of Gabriel Collins, a research fellow at Rice University’s Baker Institute for Public Affairs, home to the number one ranked energy think tank in the world.
Fortunately, we were able to help. Ursa measures crude inventories around the world (including China) using satellite radar imagery.
Radar is critical because it works regardless of weather conditions, day or night. That’s critical in a country, like China, where large cities and ports are often covered in clouds or smog.
We provided Gabe with historical data on China’s crude inventories. Armed with this information, his analysis resulted in some very interesting conclusions.
Scroll below to download a draft of the paper he wrote. Click here to access the final version on the Oil & Gas Journal website (subscription required).
Gabe focused on China’s management of key pipelines delivering oil from Myanmar, Kazakhstan & Russia and the associated inventories.
This area has been a black-box since March 2018 when the Chinese government’s statistics agency stopped reporting import data by customs entry point.
As for oil storage, Beijing considers such information a “closely guarded state secret,” making access to Ursa’s data “revolutionary,” Gabe wrote.
How oil flows into and across China can say a lot about government policy, providing valuable insight for traders.
While the vast majority of China’s oil imports still arrive by tanker, there has been a “political emphasis” on overland import routes, Gabe said.
The map below shows the pipelines and storage sites he analyzed.

Source: Baker Institute China Oil Map, GADM, Ursa Space Systems, Author’s Analysis
Gabe’s research aimed to figure out the purpose of these pipelines & storage facilities.
Is it to guard against typical hiccups, like repairs, or something more severe, such as loss of supply due to natural disasters or geopolitical events?
Is there anything to suggest these sites are being managed to maximize profit from movements in the oil futures curve?
Want to know his conclusions? Download Gabe’s paper using the link below. But first, here’s a Q&A (edited and condensed) with the paper’s author.
Ursa: You have an interesting background. You studied Chinese, lived in Moscow, worked for the Department of Defense and a hedge fund, and then became an attorney before joining the Baker Institute. When did you become interested in satellite imagery?
Gabe Collins: The satellite bug first bit me in 2008 when I tried to calculate oil storage in the Shanghai area by manually counting tanks on Google Earth.
After this initial “John Henry” experience, I came back to satellite data in 2012 while doing consulting work during law school for a hedge fund trying to get a better handle on China’s oil sector. But the data still wasn’t affordable then.
Those two research experiences alerted me to the potential value of satellite data for large-scale analysis of oil sector activity in China.
Ursa has cracked this nut, with broad global coverage data that are increasingly affordable. Equally important, Ursa’s data is especially useful for a China-focused analyst like me.
Then, in the last year and a half, China clamped down on the data that is released publicly. Bloomberg used to report weekly refinery runs, as well as monthly data on imports by port. This could be used to reverse-engineer pipeline flows.
What drew me to Ursa’s data is that China is not a blue sky environment. You need something to penetrate the fog and pollution that impair EO [electro-optical].
What were you trying to figure out here?
Where is the oil stored, how is it managed and what does this tell you about the oil flow patterns? It can also give you hints about where future infrastructure is likely to be constructed.
China is large, but Ursa data covers the ‘hot zones’ for storage. Not surprisingly, these hot zones coincide with the most active places for refining and the largest demand centers for refined products.
We still can’t see inside China’s oil pipelines, but Ursa’s data brings us much closer. You can build a flow model with a degree of analytical fidelity and insight that you couldn’t have ten years ago.
And we can now do this without endangering human sources on the ground in China who could find themselves in serious jeopardy for providing oil sector information to foreign entities.
Finally, the availability of objective satellite-measured data from firms like Ursa enables analysts to more easily audit and assess the veracity of oil inventory data periodically disclosed by official outlets in China.
Why did you begin with inland stocks?
We’re in the era of the Belt and Road Initiative and so it makes sense to examine overland oil supply routes into China and the associated oil storage, which can tell you quite a bit about current use and future intentions. These corridors can potentially transport a little over 10% of the country’s oil needs, so while a minority slice of the supply pie, they are practically and strategically meaningful.
The inland supply corridors also offer a smaller target set than China’s maritime oil import points, as well as an analytical environment in which there are fewer influences on operator behavior. This makes it more straightforward to tease out insights about how certain infrastructure assets are operated and why they are likely managed in certain ways.
Can you explain the different ways in which pipelines and inventories are managed? What’s the difference between commercial and strategic?
A pipeline being run for commercial purposes with few supply sources and off-takers would be run using the principle of stability first and foremost. You want to maintain steady flow rates. You don’t want to shut-in fields upstream or shut-off refineries downstream.
An example outside China is the TransAlaska Pipeline. There’s a concerted effort to operate within a defined set of parameters. You don’t want the North Slope to shut off or overwhelm tanks.
There are aspects of this you see with the China overland pipelines.
If you’re thinking of strategic petroleum reserves, inventory utilization rates at such facilities tend to consistently be very high.
But at Dushanzi, which is one of [China’s] named SPR sites, it looks more like a flow-assurance facility. That was an interesting insight. You would expect the utilization rate of an SPR facility to be very high and stay high.
Do you see much evidence of speculative activity?
It doesn’t seem so. With the pipeline-based sites, flows and storage facility utilization are not very responsive to short-term oil price movements. There’s been a downward trend in utilization and capacity hasn’t increased.
Furthermore, existing academic research suggests China’s state oil companies are less likely than certain international firms to use their tank storage and pipeline systems for speculative trading.
Opportunistic buying likely occurs more frequently at the coastal storage sites. These locations can take seaborne crude from a variety of global suppliers, have a wealth of local refinery customers to supply oil to, and can even ship oil back out to other regional consumers if market conditions become favorable to do so.
How much growth do you think there will be in favor of pipeline imports?
China’s fundamental oil import supply will come by sea, although there have been concerted efforts to build overland projects and increase domestic E&P operations.
Those [E&P] efforts haven’t panned out, and production keeps declining. China is trying to offset its maritime dependency, but unless there are extreme changes, the seaborne oil supply channel will be the dominant one into China. Pipelines will be the smaller pieces of the pie.