Financial Headwinds, AI, and Black Swans
Many financial analysts are pointing out that more than $120 billion in oil patch debt, with possible attendant bankruptcies, comes due between 2020 and 2022.
Again and again, news articles mention the belief that mergers and acquisitions will inevitably consolidate the industry.
What word summarizes where we are? Uncertainty.
In pricing, demand, and technology.
What tools are we as an industry deploying to reduce this uncertainty?
We cannot control worldwide demand, and because the outlook for demand is so clouded by competing forces, pricing is obviously all over the map.
We can, to some degree, control supply. Given the theme du jour on Wall Street of rigorous capital spending control, we can and have continued to ratchet down costs. As we slow the pace of development drilling in “maturing” unconventional plays like the Permian, we are seeing less volatility in WTI pricing.
Regarding technology, where are we headed?
Over the past five years we have been exposed to more and more big data analytics solutions. One of the most discussed methods of extracting knowledge from “big data” is artificial intelligence (AI).
The promise is that it will help us understand correlations between unlikely variables and help chart a course to better returns on investment. By processing huge amounts of information, we have faith that we can find the drivers that control best practices in all unconventional plays.
My problem with placing a huge amount of faith in AI is that it looks at past practices for insight to be deployed against currently accepted development models. In other words, it looks at the present to guide the near-term future based on the assumption the models of the future will be more or less the same as the models of today.
By its nature, it is blind to black swan events, those pivotal moments when everything we thought we knew changes radically.
They can be positive—like the new understanding of the tremendous potential of unconventional reservoirs. They can be negative—like the cascading panic that took down Wall street in 2008, and 2009 after Lehman’s collapse, or the recent attack on Saudi oil infrastructure.
We can’t plan for them, but we should try to do so, especially in a commodity industry like ours that invests enormous amounts of capital that generates returns on investment in years not months.
The goal of this work would be to find what’s missing in our thinking about models for the future.
For example, what work is being done to isolate and analyze the precursor events that laid the groundwork for tipping points like plate tectonics or unconventional reservoir development?
What research is being done to determine how widely—or narrowly—we should cast our data nets to find the key variables that predict coming change?
I’m no AI expert, but what I would be looking for are AI algorithms linked to predictive models of foundational variables in our industry such as:
- Supply chain volatility for completion products (water, frac sand, and rigs)
- Earth models—What are the biggest weaknesses in our current earth models and how do we resolve them? Are they growing in complexity, or simplifying?
- What conditions predict the reintroduction of conventional exploration as a larger part of operators’ exploration and development portfolios?
- Have or will geoscience and engineering inputs achieve a point of diminishing returns?
What assumptions guide us as we plan for the future?
- That the cyclical nature of boom and bust in the industry is a given
- The U.S. will continue to be at the forefront of reserves additions
- Our earth models are converging to known distributions of variables
- That our domestic demand is inelastic even if a large-scale infrastructure retrofit is enacted by Congress
- Oil & gas will be consumed in traditional ways
- Unconventional reservoir development will be the preferred target of CAPEX
1. Cyclical behavior of pricing: Are we forever locked into boom and bust cycles? Post WWII we saw 28 years of relatively stable, predictable pricing, as we did from 1986–1999. If, for the foreseeable future, production from unconventional reservoirs continues to rise, or stays at or near current levels, we could easily see oil & gas prices rangebound for a decade or more. There may be no bust, but there may be no boom either.
2. As for the U.S. being at the forefront of reserves additions, we should not forget the rest of the world is massively underexplored, both for conventional and unconventional reservoirs. By my count, the number of wells drilled in the U.S. is six times what has been drilled in the rest of the world. Most recent estimates of recoverable reserves in Guyana, Cyprus, Oman, and the eastern Mediterranean, along with focused development in Brazil and Argentina’s Vaca Muerta, should caution us that in the future the U.S. will be sharing the spotlight with the rest of the world, especially with respect to deep water targets. However, American dominance is predicated on the massive infrastructure, private ownership of minerals, and access to capital that allow U.S. oil & gas exploration to play out in a competitive market that creates an expanding knowledge base. Until other sovereign nations can embed the market dynamics that we have—and work quickly to accelerate their adoption with a stated goal of achieving long-term revenue growth—they will lag U.S. performance. An example of governments waking up to that fact is Egypt and its efforts to accelerate exploration and development of offshore natural gas in its waters.
3. Our earth models are converging to known distributions of variables: Petroleum system models fundamentally changed with the advent of unconventional reservoir development. As more wells get drilled, more LWD data is amassed; more cuttings, cores, and sidewalls samples are collected; and we’re getting more and more data about pore fluids. Are we complacent by assuming that we are refining our known models rather than laying the groundwork for entirely new reservoir distributions and depositional processes?
4. Our domestic demand is inelastic: The U.S. currently consumes about one-quarter of the earth’s daily energy output. What does U.S. consumption of oil and natural gas look like after an infrastructure overhaul? It will of course depend on the politics inherent in the funding. In a new smart grid, internet of things, 5G-enabled world, it’s a reasonable assumption that consumption patterns will change. If they drop by just 5% it would effectively take out 2 million BBL of world daily demand.
5. Oil & gas will be consumed in traditional ways: Probably, but with innovation overprints. Suncor just announced a $1.4 billion oil sand cogeneration project to replace coke-fired boilers with natural gas cogeneration to produce 800 milliwatts of power for the steam injection of its heavy oil sand project near Fort McMurray. To quote Suncor President and CEO Mark Little from a recent news release: “This project generates economic value for Suncor shareholders and provides baseload, low-carbon power equivalent to displacing 550,000 cars from the road, approximately 15% of vehicles currently in the province of Alberta.”
The news release said the project, “will reduce GHG emissions associated with steam production at Base Plant by approximately 25%. It is also expected to reduce sulfur dioxide and nitrogen oxide emissions by approximately 45% and 15% respectively.”
Note that Suncor is careful to illuminate cost and emissions savings in their statement.
Some utilities are moving to battery power packs to replace natural gas-fired peak generation facilities to meet intermittent power peak demand loads.
It would not take a huge stretch of the imagination to envision companies with natural gas pipelines tapping those for cogeneration to provide power for EV charging stations along interstates and major highways.
6. Are we at a tipping point regarding unconventional reservoir development? Costs have gone down, but this has put pressure on the critical oilfield services industry.
Bankruptcies are, unfortunately, slowly rising as companies struggle with the high-cost burdens of unconventional development. With one-month of fiscal 2019 ahead of us, bankruptcies in the universe of shale producers are 93% of all oil patch bankruptcies filed in 2018.
Might these stresses mean increasing deployment of CAPEX into conventional exploration and development?
It might, and maybe it should.
I did some napkin economics that compared a good conventional field—Covenant field in Utah—with a hypothetical Permian 60 well unconventional program.
Covenant has a projected estimated ultimate recovery (EUR) of 40–50 million barrels of oil from 38 wells. One of its better wells is, after more than 13 years, still producing about 50% of its peak oil rate—about 340 barrels of oil per day (BOPD), which was about 20,000–25,000 barrels of oil per month (BOPM).
A 60-well Permian program would need to have all wells’ EUR at about 720,000 BO per well, and cost no more than $2 million per well, to attain roughly the same economic returns.
I would point out, again, that the mild decline and longevity of Covenant’s production provides something of a hedge against pricing volatility, since it is producing through periods of both low prices and high prices.
If I managed a large acreage asset in an unconventional play, I would be tasking a significant part of the resources within my company to identify the viable conventional drilling targets within my acreage.
So, yes, we are in uncertain times.
How can we as an industry attempt to chart our future amid these kinds of uncertainties?
By thinking way outside the box and joining forces with macroeconomic, public policy, environmental, and software experts to begin to develop real-time models of the economic, political, and foreign policy drivers of oil & gas pricing, because our futures as geoscience professionals are controlled by pricing.
To get some perspectives on the kind of thinking that others have done, check out this link to the opinions of Mark P. Mills, fellow at the Manhattan Institute.