A few weeks ago, energy supermajor BP made a US$5 million investment in Belmont Technology, a Houston-based artificial intelligence (AI) start-up. The amount is small, but the potential is big. Phrases like AI, machine learning, cognitive computing and neural networks may still be lesser known in the energy industry, but they are about to get more important. A lot more important.
In the investment in Belmont Technology, BP is hoping to apply its AI cloud-based geoscience platform for more accurate lifecycle projections in exploration and reservoir modelling. Skills and analysis that used to be crunched by brute force or manually onsite are now taken to the cloud the power of new server technology and major advances in computing. By feeding historical data on geology, geophysics, reservoir and assets, the AI can throw a bombardment of computing power at the data to map it in new ways, identifying connections and trends that might miss the human eye of an analyst. Reacting to real-time data, the dynamic nature of AI technology like this allows for a 90% time reduction in data collection, interpretation and simulation.
It isn’t the only AI venture that BP is in. The British firm already has a tie-up with start-up Beyond Limits, using software used to model deep space exploration missions to refine its economic models for reservoir development, crude oil refining and fuel distributions. And it certainly isn’t the only supermajor in the area. Shell has its own AI and machine learning department in-house, tackling issues as varied as equipment failure to electric vehicle charging optimisation. Total is following in those footsteps, partnering with Tata Consultancy to build a digital innovation centre to explore AI, automation and agile methodology technology. Chevron and ExxonMobil have looked outward and building partnerships. Chevron has a 7-year deal with Microsoft to use Azure to accelerate its analytics and Internet of Things capabilities, while ExxonMobil has also partnered with Microsoft to boost its cloud technology in the Permian – aiming to improve capital efficiency and support output growth of 50,000 b/d by 2025.
The trend isn’t just limited to the supermajors. All across the entire value chain of the oil industry, players big and small are embracing digitisation. Last year, industrial specialist Rockwell Automation and services firm Schlumberger formed a joint venture called Sensia, a firm focused on ‘selling equipment and services to advance digital technology and automation in the oilfield’. In a nutshell, Sensia is promising a future where drilling rigs can run on automated schedules, oilfield equipment can communicate between themselves and machinery can assess when maintenance is required. It is making drilling for oil smarter, cheaper and more efficient with less labour. All across the world, data science-focused start-ups like VROC AI and Element AI are popping up – leveraging neural networks and artificial intelligence technology to offer novel and innovative predictive solutions to the oil industry, from more accurate reservoir management to remapping entire geological formations.
Google, Amazon and Microsoft are the leaders in large-scale machine learning, backed by their vast cloud servers – which is already powering operations at supermajors and state oil firms like Equinor and Petronas. But innovation doesn’t just lie in the very big, which is why BP’s investment into Belmont Technology is important. Belmont Technology today could be the Amazon Web Services Oil & Gas division of tomorrow. By 2035, consultancy McKinsey estimates that data analytics and robotics could save the energy industry some US$290-390 billion in annual productivity. For an industry so sensitive to costs, that’s a tantalising dream and the reason to invest now.
Infographic: Recent Big Data moves by Big Oil
Need help in getting your head around the coming onslaught of Big Data technologies?
Check out the training courses below:
Something interesting to share?
Join NrgEdge and create your own NrgBuzz today
This winter, natural gas prices have been at their lowest levels in decades. On Monday, February 10, the near-month natural gas futures price at the New York Mercantile Exchange (NYMEX) closed at $1.77 per million British thermal units (MMBtu). This price was the lowest February closing price for the near-month contract since at least 2001, in real terms, and the lowest near-month futures price in any month since March 8, 2016, according to Bloomberg, L.P. and FRED data.
In addition, according to Natural Gas Intelligence data, the daily spot price at the Henry Hub national benchmark was $1.81/MMBtu on February 10, 2020, the lowest price in real terms since March 9, 2016. Henry Hub spot prices have ranged between $1.81/MMBtu and $2.84/MMBtu this winter heating season (since November 1, 2019), generally because relatively warm winter weather has reduced demand for natural gas for heating. Natural gas production growth has outpaced demand growth, reducing the need to withdraw natural gas from underground storage.
Dry natural gas production in January 2020 averaged about 95.0 billion cubic feet per day (Bcf/d), according to IHS Markit data. IHS Markit also estimates that in January 2020 the United States saw the third-highest monthly U.S. natural gas production on record, down slightly from the previous two months.
IHS Markit estimates that U.S. natural gas consumption by residential, commercial, industrial, and electric power sectors averaged 96 Bcf/d for January, which was about 4.4 Bcf/d less than the average for January 2019, largely because of decreases in residential and commercial consumption as a result of warmer temperatures.
However, IHS Markit estimates that overall consumption of natural gas (including feed gas to liquefied natural gas (LNG) export facilities, pipeline fuel losses, and net exports by pipeline to Mexico) averaged about 117.5 Bcf/d in January 2020, an increase of about 0.2 Bcf/d from last year. This overall increase is largely a result of an almost doubling of LNG feed gas to about 8.5 Bcf/d.
Because supply growth has outpaced demand growth, less natural gas has been withdrawn from storage withdrawals this winter. Despite starting the 2019–20 heating season with the third-lowest level of natural gas inventory since 2009, by January 17, 2020, working natural gas inventories reached relatively high levels for mid-winter. The U.S. Energy Information Administration’s (EIA) data on natural gas inventories for the Lower 48 states as of February 7, 2020, reflect a 215 Bcf surplus to the five-year average. In EIA’s latest short-term forecast, more natural gas remains in storage levels than the previous five-year average through the remainder of the winter.
According to the National Oceanic and Atmospheric Administration (NOAA), January 2020 was the fifth-warmest in its 126-year climate record. Heating degree days (HDDs), a temperature-based metric for heating demand, have been relatively low this winter, which is consistent with a warmer winter. During some weeks in late December and early January, the United States saw 25% to 30% fewer HDDs than the 30-year average. This winter, through February 8, residential natural gas customers in the United States have seen 11% fewer HDDs than the 30-year average.
Source: U.S. Energy Information Administration, based on National Oceanic and Atmospheric Administration Climate Prediction Center data
Headline crude prices for the week beginning 10 February 2020 – Brent: US$53/b; WTI: US$49/b
Headlines of the week
Global liquid fuels
Electricity, coal, renewables, and emissions