Unfortunately, Internet Explorer is an outdated browser. Due to security vulnerabilities we do not support it. To have the best browsing experience, please upgrade to Microsoft Edge, Google Chrome, Firefox, Opera, or Safari.
The chemical, oil & gas and the process industry as a whole are in the midst of a digital transformation. Many companies have drafted strategic digitization roadmaps in which various aspects of their businesses are transformed and digitized. For many companies, digitization of operations has two main goals, (1) to eliminate paper trails or undocumented processes and (2) to capture and process more data and information than the company had ever deemed possible in the past.
Your digitization strategy has two main goals
The expectations from eliminating paper trails or undocumented processes are quite straightforward; it is about creating additional documented traces of activities and events for future reference and preventing people to spend time on paperwork. Think for instance about the entire process of work permits in which different actors such as maintenance managers, process operator and subcontractors need to sign off all activities before any work can be executed in the field for safety purposes. Manually signing-off these permits is very time consuming and tracing these papers is a daunting task. Digitizing this process with (online) applications and field tablets has the immediate benefit of faster and more efficient communication streams and quickly searchable trails and records. Alternatively, take the example of field inspections or valve operations. In many cases these are manual processes in which an operator checks a pressure meter, oil level gauge or valve position in the field and reports back to the control room where the inspection is documented on paper. Not only does this take time, it is also very prone to human error or negligence. In this case, digitization could mean two things, or the operator is given a tablet in which the inspection can be directly performed in the field and stored in a central system, or the manual inspection in its entirety is replaced with a sensor which autonomously and continuously performs the inspection, not only eliminating human error, but also offering reliable real-time data and alerts when something is about to go wrong.
This last example touches upon the second expectation from digitization, to capture more information than the company had deemed possible in the past. Although many plants and facilities are already heavily automated, still many relevant information is not yet captured. For instance, to monitor emissions in a plant, a handful of VOC sensors are deployed in critical locations across the plant. Increasing the amount of sensors would give a much more precise view on the levels and locations of emissions and potentially allowing to identify leaks in realtime. Take as a second example corrosion of piping under isolation, which is not visible from the outside. Common practice today is to do regular inspections in critical locations or just act upon problems when they occur. When deploying corrosion sensors on every pipe, plant operators and maintenance managers could have a real-time view on the state of each pipe and prevent problems even before they materialize. These two examples are only scratching the surface of the potential of digitization by means of sensors. Many more use-cases are top-of-mind: monitoring the state of each manual valve across the plant, signaling pops of pressure relief valves, alerting on malfunctioning steam traps, monitoring abnormal vibrations on every unit of rotating equipment, tracking the temperature and pressure within each segment of piping, etc.
Many digitization managers dream of capturing every possible piece of relevant data. Digital operations are high on the corporate agenda. Indeed, a recent study by PWC shows that over 40% of CEOs of chemical companies will invest in digital operations in the coming 12 months.
The main barrier however today is that the sensors to capture all this information are not yet available, or if available, too expensive to be deployed on a large scale. The cost of deploying current generation instrumentation equipment is mainly driven by the engineering time and effort required to install the cabling towards all these sensors, which in many cases has a negative impact on the individual business case of deploying the solution at large scale. In this context, the Industrial Internet of Things offers opportunities to drastically rethink the business cases of these technologies.
Industrial Internet of Things is an important value driver
The Industrial Internet of Things made the promise to support the digitization objectives of industrial companies with relatively cheap, wireless and battery operated sensors and actuators. Depending on the type of application or use-case, various objectives are being put forward:
Improve efficiency of plant operations and maintenance activities such as reducing the time spent by operators in the field, eliminating administration and capturing new data to make better and faster decisions.
Prevent unplanned down-time of the plant by monitoring previously unmonitored plant parameters and capture data that can assist the maintenance team in predictive maintenance actions.
Increase safety of plant operators and reduce risks that might lead to incidents, accidents and environmental spills.
The ultimate goal of the Industrial Internet of Things is creating a digital twin of the entire plant. In a digital twin, every parameter of the plant is digitized in such a way that one could run a digital simulation next to the physical plant and predict the behavior of the physical plant by changing some parameters in its digital environment. A digital twin gives a real-time view on the current operating conditions of the plant and presents an intelligent tool to manage change and reduce risk.
The hidden gem of a digital twin is in its data. Once the operating conditions of the entire plant are digitized in real-time, i.e., every valve position is known, the temperature, pressure and flow within every meter of piping is registered, the vibration levels of every bearing are monitored, every leaking steam trap is recorded, etc., all data is available to accurately predict the behavior of the plant. Obviously, it is impossible for operators to do this manually, hence, advanced machine learning and artificial intelligence could be leveraged to assist them doing their job. For instance, a machine learning algorithm could learn the normal operating conditions of a plant over time and give the operator an alert when it notices an anomaly in the current state, e.g., the pressure within a pipe raises too high in combination with the rotation speed of a ventilator unit, or a set of manual valves is operated in an order not yet observed in the past.
Building your business case
Before spending any money on implementing new infrastructure, any company will perform a cost – benefit analysis of its investment. Let’s take a moment to look at some common business case for IIoT projects.
Projects focusing on improving the efficiency of operations are quite straightforward. Typically, there is a clear problem such as “operators spend too much time on inspection rounds”, which is clearly quantifiable such as 8 hours per week. By installing sensors to do the inspections in real-time, the plant could free up a number of hours per week of the operator time to spend on more productive tasks. Calculating the return-on-investment can be predicted pretty accurately.
Prevent unplanned down-time is somewhat more complicated as you need to be able to predict the chance of the plant or a process going down due to a cause that can be solved by an IIoT implementation. Once you are confident about this chance, the cost associated with such a down-time can be calculated with quite some accuracy (production loss, intervention costs, cleaning contaminations, etc.) and the business case can be developed further.
Projects aiming to increase the safety and reduce environmental impact (spills or contaminations) are the hardest to build a business case for. First one needs to be able to predict the chance and magnitude of such an incident or accident and second, one needs to determine the “cost” associated to this. While some companies try to build this business case by putting a cost on an accident, many go with the saying “If you think health and safety is expensive, try an accident”.
The projects described above are also simplified quite a lot, as in practice, many projects are a combination of many different objectives.
How to unlock the full potential of your digital strategy
As stated in the introduction, many companies already have drafted strategic digitization roadmaps, including the Industrial Internet of Things as an enabling technology. Typically, such a roadmap is structured as a list of projects, each having an individual business case as illustrated previously with a timeframe associated to it. Individual projects are identified bottom-up and a business case is made on a per-project level. The trap of this approach is to only focus on point solutions with the best individual business case which will unlikely unlock the full potential of a digital twin. So how do we get from isolated use-cases to a digital twin? Some recommendations:
Picture the ideal implementation of your digital operations facility. Take a holistic perspective covering both regular operations and maintenance activities. Determine which operational data is relevant and could be a value driver and which data is not relevant?
Think big data. Many digitization projects focus on the data and information which is relevant today, which boils down to the information that can be processed by operators or analysts, neglecting the (future) opportunities of artificial intelligence or machine learning. Take the example of manual valve monitoring. Monitoring the state of a select number of critical manual valves will immediately return benefit as it can be processed by an operator in the control room and eliminate part of the incidents. However, to eliminate all incidents caused by incorrect valve positions, the sensor data will need to be processed by algorithms to autonomously signal potential risks and hazardous situations.
Start with low-hanging fruit but expand quickly. Take the ideal implementation of your digital transformation as a guiding point to scale up your individual projects. It is impossible to draft a business case for the entire digital transformation, so prioritize the low-hanging fruit, i.e., a select number of projects with a clear short-term and promising long-term return. Before engaging in numerous other small short-term projects and losing focus, expand the scope of these initial projects to capture the long-term return, for instance scale up from an individual plant to a site-wide rollout.
Take joint infrastructure out of the equation. Industrial IoT projects usually require infrastructure such as wireless networks and server infrastructure that can be shared with other business units within the company. The cost of this shared infrastructure should not be attributed to the business case of the first IoT project that is implemented. We recommend taking this out of the individual business case or only account a fraction of the cost.
Leverage the cloud for your operations. As the amount of data that needs to be processed will keep increasing it will get very complex and expensive to process this on-site. Public or private cloud processing facilities offer a much more scalable and maintainable solution to the challenge; however, this might require changes in company-specific policies on how data is treated, and which data can and may not leave a geographic location.
Rethink your organization. Many companies are subdivided in IT (information technology) and OT (operational technology) departments. In the world of IoT the responsibilities of these departments are starting to overlap. To make a digital transformation work, it is important to involve both departments and start thinking about how to structure this in the future.
Rolling out your digital transformation roadmap and implementing IIoT solutions will take time. It is a gradual approach in which the solution will grow over months and years. While many plants will never manage to build a full digital twin, keep in mind there is already a lot of value to capture along the way.