A must-read new book on the oil and gas industry

The volume of data our societies create, represented as zeros and ones in a storage device, has been growing by truly staggering amounts. According to the International Energy Agency (IEA), IBM estimates that between 2015 and 2016, the world generated almost as much data (90 percent) as already existed in all of the world’s storage systems. Our devices and our increasingly digital lives contribute to this growth rate in many ways.

● A high-quality photo comprises eight to ten megabytes of data. Most of us do not even discard photos anymore. They simply pile up on our phones and home computers.

● A high-quality ten-minute video taken on our smartphone takes up 1.5 gigabytes. Four hundred hours of video are uploaded every minute to YouTube.

● A typical flight produces a terabyte of data, and autonomous vehicles and trucks will generate similar data volumes.

Industrial data volumes are not growing at quite the same pace as consumer data, but that’s because the industry has not yet equipped all its various assets, tools, and people with sensors. But as industrial assets become data generators, they will match the prolific data generation of humans. An Airbus or Boeing aircraft is festooned with hundreds of sensors that produce discrete data measurements every tenth, hundredth, or thousandth of a second throughout a flight.

Beyond the growth in volume, data is also changing shape. Early generations of computer systems could only process highly structured data, such as rows and columns on a spreadsheet, in the form of numbers and letters. But modern data can take almost any shape, including unstructured data like photographs, waves, sounds, video, and sensations like vibrations and smells.

The DNA of living things can be represented as data, a very important development. The industry will soon be able to tell the provenance of a thing or substance by investigating its deoxyribonucleic acid (DNA) and comparing it to a registry of DNA samples. The oil and gas industry is not far from a future where it will be able to take a sample of a barrel of oil and identify which oil basin it came from by testing microbes found in oil samples (although perhaps not after crude oils are blended in pipelines and tanks). Ethical producers of oil, those nations that invest more to protect the environment, will be able to command a premium for their product.

An emerging challenge is how to analyze this flood of data. Tools and techniques for data analysis, interpretation, visualization, and monitoring need to keep pace with the growth in data volumes, the flow of that data from one location to another, and the analytic demands of data users. The spreadsheets of the past are simply not up to the task.

Analytics, a term for computer horsepower, is also demonstrating the same rapid development and growth as data, which isn’t surprising because analytics are carried out on computer chips, too (though not the same chips that store data). Think of “analytics” as consisting of these math computer chips mounted on circuit boards and the software that carries out the math. Both data and math chips have been advancing with vigor, and along important lines:

● Chips can be both highly specialized (for such intense tasks as video gaming or robot controlling) and generalized (for laptop and desktop computers). Car makers now use video-gaming chips’ ability to process visual data as the basis for autonomous car navigation systems.

● Sensors are basically chips. Inexpensive sensors for GPS, orientation, cameras, light, sound, and speech-processing are all just math chips with embedded software.

● Chips need power to operate—to keep them running longer and more continuously. To prevent them from overheating and depleting their batteries, engineers have been designing them to be power-stingy, and power consumption has been falling with each generation of chip.

● Along with their power profile, physical size, and capabilities, their costs have fallen dramatically, allowing chips to be installed in many unexpected devices including pumps, valves, and gauges.

We can see the impacts of these developments on a personal level. My Apple Watch, which retails for a few hundred dollars, has the equivalent analytic capability to a multi-million-dollar Cray supercomputer from the 1980s. A smartphone has much the same capability of the mainframe systems that enabled NASA’s moon-shot programs in the 1960s, and, in just ten years, smartphones advanced from a novelty to a must-have for modern life.

Industries have varied in their drive to put sensors onto things, but the pace of adoption is accelerating. Roland Berger estimates that 15 billion smart sensors were installed globally in 2015, but by 2020, the number of sensors is expected to be greater than 30 billion. This phenomenon is giving rise to the Industrial Internet of Things (IIoT)—pumps and motors, valves and gauges, pipes and tanks, wires and switches are being fitted and retrofitted with sets of sensors. Once in place and live, the sensor adds to the flood of data, draws on the power grid, and creates demand for connectivity to send that data elsewhere.

The prize for the industry is enormous. Greater visibility through sensors means that health and safety issues, like leaks and spills, can be detected faster, even before they occur. A better understanding of equipment performance means equipment can run longer and at higher rates without failure. Faster collection and processing of more accurate field data means faster payment by oil companies for the services they purchase.

Chips that can perform calculations need software code to execute the math. Similar to the chips, software development shows the same exponential growth characteristics.

● Programming languages are becoming easier to learn. They feature drag-and-drop interfaces, high reusability, and standardization so that coders are more productive. Compute needs like sorting, analysis, and mapping are simple and built-in.

● Programming languages are becoming more ubiquitous and common to multiple uses, rather than specialized by asset or application. Recruiting coders is much easier.

● The use of common chip sets means the development of industrial software for a piece of equipment is not dramatically different from coding for a commercial system or a web application.

● It is now acceptable for companies to rely on open-source code software that is developed freely and shared openly—which both accelerates the delivery of new solutions to business and lowers the cost.

The final building block in a digital world is connectivity. Without connectivity, a device that has data and analytics is just a calculator and, these days, not a very useful calculator.

Low-cost chips, analytics, and software development have helped transform the telecommunications sector in just one human lifetime to enable extraordinary connectivity. From 1991, with the launch of 2G, the telecom industry has progressed to 3G, 4G, LTE, NFC, Bluetooth, and now 5G.

As a result, the world is becoming highly interconnected. According to the IEA, at the dawn of 2016:

● The number of households globally with Internet service was around 54 percent, compared to 80 percent that had access to electricity, concentrated among developed nations, but accelerating in the developing world.

● The number of individual Internet users was about 3.5 billion.

● The number of mobile phone subscriptions, a measure of the number of users able to tap into digital services, reached 7.7 billion.

● The number of inanimate objects or things that are connected to the Internet is estimated to be about 8 billion and will grow to 30 billion by 2030.

The volume of data that networks move provides a good indication of the penetration of and demand for connectivity. In 1974, the total amount of data that was transmitted on worldwide networks in a month was one terabyte. By 2016, worldwide networks move one terabyte every second, an increase of 2.5 million times.

There are no signs that this volume of data is flattening out. If anything, the growth in the number of sensors and the greater capability of analytics suggests that data volumes are likely to continue to grow. Indeed, certain technical innovations in our digital world are still very early in their own adoption life cycle (such as autonomous vehicles), suggesting that the demand for data connectivity has serious propulsion.

One clear issue for the oil and gas industry, even in very developed nations like Canada, and still very pressing in developing nations, is the lack of robust and ubiquitous telecommunications infrastructure. Oil and gas is often found far from concentrated areas of civilization, and telecom companies have been slow to roll out network connectivity. This continues to plague both Australia and Canada, as well as the offshore industry in general. Telecom technology still suffers from intermittency during weather events, a serious challenge for dangerous infrastructure that needs constant supervision.