What is the Future Demand for Petroleum Engineers?
Artificial intelligence can do some things faster and better than humans can. But that may allow engineers to focus on higher-value tasks.
At the end of a presentation on BP’s annual energy outlook last year, a petroleum engineering student from Greece asked, “Is a job as a petroleum engineer a good idea for the next decade?”
Then BP CEO Bob Dudley responded by saying, “A job or training as an engineer of any kind is so important. There is going to be a huge need for that as you can see from Spencer’s outlook.”
He was referring to the presentation made by Spencer Dale, group chief economist for BP, whose outlook showed oil demand likely growing until 2030, and remaining around 100 million B/D through 2050 while natural gas demand continues to rise.
But they never said how many oilfield engineers would be working in the decades ahead. A JPT online article about that exchange was one of the most read of 2020, suggesting it was a question on the minds of many readers.
The most downloaded SPE paper in 2019 on OnePetro was “The End of Petroleum Engineering as We Know It,” which predicted a significant change in what engineers do and the number needed to do it over the long term (SPE 194746).
The idea for the paper came when 2016 SPE President Nathan Meehan said he began to have doubts about whether the assurances he was offering when asked that question by students were a realistic appraisal of the future.
After oil prices began to crash in late 2014, UK oil and services companies slashed their payrolls. By 2018, they employed 38% fewer workers, according to Oil & Gas UK. The oil industry group said that during those years, the industry reduced the cost of operations in the North Sea basin—the world’s most expensive by far—and production began slowly rising again.
In the US, oil production rose 46% from 2014 to 2018 while the number of petroleum engineers working declined 3%, according to the US Bureau of Labor Statistics. There have been more job reductions in the US since.
Those results show how the number of workers and oil production can move in different directions over a short period of time. Students asked BP’s top executive a tougher question during the energy outlook presentation, such as, will technical professionals hired now be needed in 20 years?
BP’s outlook did not estimate the number of jobs needed in the long term, nor did the SPE paper on the future of engineering. Any predictions would require some big assumptions on variables, such as future oil demand.
For example, oil production in 2040 could range from 80–130 million BOE/D, according to BP’s outlook. The low estimate assumes a politically difficult shift to policies requiring rapid emission reductions. The high estimate assumes strong demand growth. Two other scenarios put demand around 100 million BOE/D, with minimal growth after 2030.
Technical staff levels will vary based on the demand. At the low production level, oil companies will need to find 4.5 million BOE/D worth of production each year as older fields decline. If production is at the high end, that total will rise to 7 million BOE/D.
The payoff for digital change is also hard to predict. Computers have been around for decades with little impact on staffing. That may change but the industry is still in the early stages of a digital shift.
Investor valuations put a lot of weight on an oil and gas company’s growth and profits.
Slow-growth majors stick to high dividends, even when times are tough, because cutting their dividend is likely to hurt their stock price. Generating the free cash flow to keep those dividends requires a disciplined focus on finding the lowest-cost, highest-value hydrocarbons.
“Some investors are questioning the need to explore at all given the vast discovered resource base yet to be developed,” said Alana Tischuk, who works with consultancy Wood Mackenzie’s global exploration team.
A commitment to production over exploration could affect the technical teams associated with it, but the upside of new discoveries would be lost.
Successful companies will be the ones “avoiding the most difficult, expensive, and risky prospects,” said Robert L. Kleinberg, adjunct senior research scholar at the Center on Global Energy Policy of Columbia University, in an article in The New York Times.
BP’s outlook concluded that curtailing oil and gas exploration would result in production levels insufficient to meet even its lowest demand scenario.
Tischuk predicted that companies “will drill in the hope of finding better resources than those they already have—lower-cost barrels with a higher margin.”
One analogy for the future of exploration and production is coal mining. Both are highly cyclic sources of cheap hydrocarbons from the ground and both produce fuel that emits carbon dioxide when burned, which scientists say is leading to climate change.
Coal emits more CO2 and other pollutants than alternatives such as natural gas. Oil companies have used that fact to position gas as a cleaner substitute for coal-fired electric generating plants.
The SPE paper noted that mining engineering has declined dramatically over the decades. It is not the only example of an industry where deep slumps have led to intense cost cutting that has reduced demand for engineers. Copper mining, textiles, and aeronautics are also on the list of companies in cyclic industries employing fewer engineers.
While petroleum engineering is down now, 6 years ago when oil was worth $100/bbl, it was at the top of the engineering jobs growth list.
Public attitudes toward the future of oil also offer conflicting signals about the future of petroleum engineering. Dudley said that while oil jobs are coveted in many countries, in the US and Europe many have a negative perception.
The many names for various digital innovations get confusing, but the potential impact on jobs is brutally simple.
“Collectively, these advanced technologies should enable one engineer to do the conventional work of many,” the SPE paper said. That power becomes useful if an engineer can find a way to use rapid-analysis tools to reduce the cost of producing a barrel of oil, or better yet, find a lot more of it.
Recently, BP reported that it used advanced seismic imaging to add low-cost production near its Gulf of Mexico platforms, saving the cost of building a new platform, and that it is also using high-tech tools to find oil in ultradeepwater.
The Digital Difference
There are tasks where a machine can do the job faster and better.
“There is no reason decline curves are done by hand. AI tools should be better than people,” said Meehan.
Other jobs on his list of automation targets include log analysis, pressure-transient analysis, and a lot of well test interpretation.
Those doing digital analysis are looking for new ways to speed up the work. Jeff Yip, a petroleum engineer for Chevron, said he often asks himself, “What am I doing that is taking up time? How can I automate this? How can I take this out of the realm of something manual, off of the to-do list of the employee, and shift it to the to-do list of a computer?”
But machines can only do so many jobs. The SPE paper estimates that only 20–25% of the jobs done by petroleum engineers now can be automated. The larger impact will be programs that augment what an engineer can do.
A machine that can constantly analyze the water injection and production trends in a large, old field can drastically reduce the hours people spending tracking what is going on, allowing a few engineers time to look for ways to eke out more barrels.
The reward for those who increase what they do digitally may be continued employment. For example, Mark Bahorich, cofounder of Q Engineering, which sells a cloud-based service that generates decline curves faster than a human does, said none of the investment banks using its software has laid off workers analyzing oil production.
“The parts of their jobs we automated allowed them to focus on higher-value tasks. Technology in this case enables engineers to solve the right problems instead of the old ones,” he said.