Oil and gas will become one of the driving forces in industrial IoT (IIoT). Oil and gas facilities generate tremendous amounts of data, and only a fraction of it gets used. Companies like Shell – which tracks over 7 million data points on systems ranging from LNG tankers to drilling equipment – utilize digital technologies and predictive maintenance methods that result in huge productivity gains, increased safety and lower costs. In short, IoT is helping automate the oil and gas industry with increased access to valuable data.
While there are success stories, there’s also a lot of misinformation and hype that make it challenging to understand what’s realistic. Below are five questions and a set of best practices that can help customers navigate these challenges as they move toward automating with IoT and predictive analytics.
-
What is IIoT
To people in the IT industry, “IoT” and “IIoT” often means collecting and analyzing machine data to achieve a business goal, such as energy consumption. To them, IoT is the new IIoT – representing an expansion and augmentation of functions they already perform. For years, they’ve been analyzing data from SCADA systems that link many of their major systems —gas plants, booster stations, electrical transformers, SCADA and DCS systems, etc.– to optimize their operations.
This is an important distinction because (1) it will help you avoid confusion when you’re talking to different parts of large companies, (2) it highlights to a company that they may already own much of the technology for IIoT and (3) it brings into focus one of the biggest challenges.
-
Where are the potential data siloes?
Simply put, it costs far less and takes less time to link new or stranded assets through wireless technologies and IoT gateways, than retrofitting existing SCADA or DCS systems. For example, connecting untethered devices with IoT gateways and stick-on sensors can cost a few thousand dollars; but upgrading a SCADA system to do the same might cost $25,000 or more. Additionally, there is a cost of dilution of the core functions of SCADA and DCS systems, which purpose is to safely control the process, not to do large amounts of data acquisition.
You have to remember to merge data sources or else you’ll end up with silos providing an incomplete picture, which inhibits automation. The best strategy is to integrate these different sources of data into an operational data management infrastructure layer that allows your employees to extract from a single, trusted source into any application. Your goal shouldn’t be to create a solution for the engineers, but instead to develop something that a broader group can use.
-
What is the business case?
Increasingly, companies want to see payback in two years or less from IIoT projects, and often the best way to achieve these results is by focusing on predictive maintenance and improved uptime.
For instance, DCP Midstream is one of the largest natural gas processing companies in North America. They have 61 gas plants, 57,000 miles of gathering pipelines, and 400 booster stations with more than 1,400 compression units spread across the middle of the country, forcing crews to travel millions of miles per year to maintain the equipment. .
Additionally, many of their booster stations and associated compression units only served up a small subset of data gathered in their SCADA, so they didn’t have the full picture of their operations.
To better manage its fleet, DCP employed an edge/IIoT sensor augmentation solution across 130 of their 400+ booster stations based on economic risk. Each compressor system that included a booster station gateway to a cloud-based analytics solution with a specialized portal for visualization was installed in less than 2 hours.
While the above is just a portion of what DCP Midstream’s plans as part of its digital transformation initiative, the use of IIoT is an important piece. In the first year, the company estimates it saved $20 to $25 million, which paid for the first year investment with an estimated incremental $20 million in 2018.
-
Who is accessing the data?
With IIoT, not only are you collecting more data, you’re creating a system that encourages more people to use the data, which requires thinking more about the user experience. Having mobile access to contextualized operational data coming from all data sources is an absolute necessity, so it’s always available to operators and maintenance technicians in the field responding to emergencies, as well as executives at the airport who want to compare current production with pricing forecasts.
User interfaces will also need to be flexible. A factory manager who wants to solve an immediate production problem will want a different view of the data than a competitive strategist who’s trying to figure out the optimal configuration for a next generation facility. The time required to obtain, understand and act on the data is one of the key factors of success. If people can’t start quickly using this data in their daily routines, your project will likely fail. Providing easy to use, contextualized data in a mobile, self-serve environment is fundamental.
-
Who is performing analytics and where are they taking place?
When you say “analytics,” people often envision data scientists leveraging algorithms to sift through mountains of data in the cloud. Some believe complex problems that are solved with IoT are better processed in the cloud, and others say the edge of your network is the better option for processing data. MOL, a large refiner in Hungary, for example, is conducting analytics on the edge with its machine data to determine how and when it can utilize less expensive, higher sulfur crudes in one of its refineries, without introducing risk to its operating parameters. For MOL, processing at the edge is more beneficial that sending it out to the cloud and all the way back.
Still, whether edge or cloud, many analytics problems are being performed directly by people looking at a few finite data streams, which could be viewed as “human analytics.”
Companies should think about taking a “layered” approach to analytics and take into consideration edge vs. cloud capabilities and costs, whether human or algorithms might work better, and the feedback loop between these systems to ensure insights gets operationalized. You can’t anticipate every use case you will have in the future, but if you take a broader view of what “analytics” means, you can create an analytical framework that can guide in the decision making process when the need arises. In general, performing real-time analytics closer to the edge in an operational data infrastructure that includes IIoT/edge produces better, more sustainable results.
