13 Mar Digital blockers, barriers and solutions
The IEA recently asked me about what I saw as the key blocker and barriers to the digitalization of the oil and gas industry. Here’s my response.
This post would be valuable for anyone thinking of selling digital solutions to an oil and gas company, and needing answers to the objections that will inevitably get posed.
The IEA is one of the oil and gas industry’s go-to sources for reliable publications on energy supply and demand, and they like to keep abreast of impending changes that might have an impact on energy markets. Digital technologies have dramatically impacted financial services, transportation and the travel sector, and it stands to reason that the oil and gas industry is going to be affected eventually. The question is how, when and by how much.
Let’s begin with the big one – the brownfield challenge.
Brownfield oil and gas infrastructure (field assets, pipeline compression plants, gas plants, refineries, upgraders) is already installed and operating, compared to greenfield, or brand new and still on the drawing board. It’s way easier and cheaper to make digital changes to greenfield infrastructure because the changes are basically to a paper design.
A brownfield plant may be energised (i.e. high voltage is flowing through it, or is heated), or is pressurized, or is rotating (i.e., pumps are active), which means it has to be switched off, cooled down, drained and devaporised before work can be done on it, including making digital changes. Shutdowns like this don’t happen on a frequent basis. They can take a year to plan and may even be years apart. It can be impossible or at the very least impractical to even test a digital adaptation in a live and running plant.
For digital innovation to experience uptake in brownfield settings, it must demonstrate a sensitivity to the challenges of brownfield plant operations and the difficulties of introducing change into live environments.
Considering the brownfield problem outlined above, imagine how hard implementing change must be in an industrial setting. Needless to say, change happens cautiously, for which we’re all grateful.
Plant managers, supervisors and staff are highly motivated and rewarded on such measures as zero safety incidents, high reliability, predictable cost and high quality. Employee turnover in plants is usually very low, creating a stable workforce, but in many instances, a strong aversion to process changethat create either perceived or real risks.
The usual approach to digital innovation (prototype, fail quickly, iterate) works best in a dynamic, agile, adaptable and flexible work climate and culture. In other words, the opposite of what you find in the standard industrial setting.
For digital innovation to succeed, it needs to introduce change slowly and patiently, with the same easy to use adoption curve as the best digital innovations anywhere, all while protecting the key performance measures.
I sense that the engineers in oil and gas tend to view what we think of as “digital” (consumer technology) as being in the realm of toys and just not robust enough for industrial use, where the environment and lives are at risk. Remember the “blue screen of death”, that curious way that Windows systems can simply freeze on you for no reason? There’s no place for this system behavior in an industrial setting. What if, just when you need to cycle a valve to address a pressure event in a gas plant, the system just shuts down or freezes, or the rotating hourglass starts a perpetual loop?
Acceptable digital technologies are those that demonstrate a level of robustness consistent with industrial standards.
Incumbent technology design
Many of the key industrial grade automation technologies for oil and gas infrastructure were not designed to be “open” in the same sense as open systems that allow interoperability between different supplier technologies. In fact, given the risks involved, it actually makes sense to be “not open” so that more control can be exercised over the performance of the technology.
Industrial technologies such as programmable logic controllers (PLCs) or SCADA systems, simply lack convenient or easy ways to allow integration with other technologies, including digital solutions.
And then there is the question of the systems architecture to consider. Say a pump is connected to a PLC that controls the pump’s actions. The pump’s various sensors feed a steady and high volume of data to the PLC (temperature, rotating speed, pressure, voltage, throughput). There’s little benefit to moving this data beyond the point where it’s generated. Engineers ask what, if any, of the torrent of data should move beyond the PLC to some other system for analytics, or reporting or monitoring? Where and how should a digital solution plug into this data flow?
This is a key systems architectural question that bedevils plant designers. They need to consider all the equipment in plant,the PLCs and their limitations, the thousands of possible sensors, the bazillion data points that the sensors generate, and the need to monitor or supervise the whole lot from inside the plant, in the plant control room, in the maintenance department, in production, and in the head office.
Good digital technologies consider their data generation and data consumption needs, and how best to fit into the architectures in place.
Unlike commercial systems, like ERP, that have been plugged into the internet since the 90’s, most plant systems are closed and not linked to the internet. They lack the kinds of cyber monitoring or industrial design thinking that keeps hackers at bay. We can see this phenomenon play out in the frequent news stories about industrial equipment (and even consumer technology) gone awry because of cyber problems.
Considering that most industrial plants were not intended to be connected to the internet (many predate the internet!), and suffer from the brownfield problem outlined earlier, the only practical solution is to prohibit the introduction of uncertain technologies into the plant setting. That means no cell phones in plants, no internet connection, no cameras, and soon, probably no toasters in the lunch room.
Digital solutions that meet the highest standards for cyber security will be more welcome.
One of the hazards of industrial settings is the possible presence of flammable or explosive vapours, some of which are invisible and hard to detect. These vapours may only need a single spark from some electrical device that’s not properly grounded to ignite a fiery conflagration. Most consumer technology falls into this category – not intrinsically safe.
Intrinsically safe devices are significantly more costly than consumer devices, having a much smaller market, low economies of scale in manufacturing, and an impressive feature set (waterproof, bombproof, crush proof, impact proof, etc).
Industrial plants can often produce very high levels of electrical interference, or because of their scale, render mobile networks unstable, unreliable or unavailable. Most plant equipment is therefore hardwired to a physical copper or fibre optic network.
Many oil and gas fields are far removed from civilisation and won’t have access to a mobile network. Unfortunately many digital technologies are designed for dense urban settings with good network coverage. Mobile devices that rely on a cloud network connection may not work.
Digital technologies need to be intrinsically safe where it’s required, and must be able to run in the absence of a reliable network connection.
Timing and economics
It’s pretty tough for a digital innovation to be the best marginal investment in oil and gas. Consider just the simple math of a producer of 100,000 bbls per day. Assume the price of oil moves from $45 to $50, or $5/bbl. That’s an addition $182m revenue in a year and no increase in cost. For most oil and gas companies, the best strategy is to not do anything and wait for the price to rise.
Effectively, the only time a digital investment would be the best marginal investment is when production costs are well above the recovered price (the net back is zero or negative), supplier costs have been slashed sufficiently to trigger bankruptcies, capital spend is just keeping the lights on, and the workforce has been trimmed to skeletal levels.
In other words, when there’s nothing left to do.
If digital investments can help lower costs further, or extend the capital dollar, or improve the productivity or safety of the remaining staff, and quickly, they will make sense.
Fortunately, we’re at precisely this moment in the commodity cycle.