While digital transformation has been getting its fair share of hype in recent years, end users in the heavy process industries actually began digitizing their plants decades ago with the introduction of “smart/intelligent,” digitally integrated process transmitters and final control devices. Today, the vast majority of process transmitters installed for greenfield and major upgrade projects are smart, if not always fully digitally integrated. However, historically, pumps and valves have tended to be the last field devices for end users to digitize in their plants.
Advanced diagnostics and bi-directional communications can help improve process performance, condition monitoring and maintenance effectiveness; while reducing maintenance costs to a significant degree. This is particularly true for the traditional “bad actors,” such as control valves. Smart valves and pumps could also help reduce fugitive emissions to improve environmental compliance, and – in the case of safety-related valves – enhance plant safety.
With the emergence of Industrial IoT (IIoT)-enabled remote management solutions, process industry end users can begin to take full advantage of remote monitoring, analysis, and management services provided by valve and pump suppliers or third-party service providers. While leveraging the expertise of these external partners would appear to be a “no-brainer,” ARC research indicates that many end users remain resistant to this concept.
Two examples of services that suppliers have developed to help their customers lower overall total cost of ownership of their valves and pumps follow. Significantly, both services can also enhance plant safety and regulatory compliance.
“Expertise As A Service” For Valve Maintenance
Several few years ago at an ARC Industry Forum in Orlando, Florida, Shawn Anderson, senior research specialist for Emerson Process Management, gave a presentation on how that company is leveraging the IIoT to help end users reduce valve-related unplanned downtime.
Anderson’s group initially began looking at adopting IIoT technologies as a way to collect more valve health data from the field and provide more realistic valve failure information than could be generated in a lab. It soon became apparent that IIoT technologies were a natural fit for developing a remote monitoring service geared at optimizing customers’ valve maintenance practices. This takes the form of a connected services work flow.
According to Anderson, what end users really want and need is actionable control valve health information. End users want to know what they need to do and when they need to do it to keep operations up and running. ARC research confirms this.
Clearly, end users can benefit from partnering with a trusted valve supplier (or third-party services provider) that can help remove the burden of valve maintenance to enable them to focus on their core competencies. Innovative new service models such as “Expertise as a Service” (EaaS) in which remote monitoring and diagnostics services are bundled provide new options for end users lacking in-house valve experts.
To succeed, this typically requires close cooperation between the valve or third-party supplier that has the valve expertise, IT suppliers that can provide the secure IIoT platform, analytics suppliers that can provide the appropriate analytics and visualization tools and the end users themselves, who must be willing and able to provide the raw process data and, ultimately, act upon the information.
Machine Learning Provides Advance Warning Of Pump Failures
At another ARC Industry Forum session, Rob Miller, general manager, Global Solutions for Flowserve, presented an example of how that company is implementing new maintenance and reliability practices for its valve customers by integrating advanced machine learning technologies. Flowserve has been evaluating machine learning capabilities to improve its equipment monitoring capabilities for the past 20 years, but the results had proven to be too costly and difficult to commercialize. However, technological advancements in data science, artificial intelligence and computing power developed in the last five years have changed that paradigm.
As we learned, for nearly a decade, Flowserve had been eager to increase its asset performance and advanced diagnostics capabilities for its product portfolio. The initial approach was to increase data acquisition capabilities utilizing wireless technologies to bring pump health analysis data up to the plant or cloud level, and eventually migrate to actively monitoring customer equipment. However, the company faced many challenges related to predicting equipment failures with adequate advance notice to allow its customers to react effectively to alerts.
Recent technological advancements have enabled Flowserve to refine its methodology and introduce a step change in its equipment monitoring capabilities. Integration of machine learning techniques and cognitive analytics have enabled the company to provide more advance warning of impending pump failure to give customers enough time to prevent the failures before unplanned downtime could occur. According to Mr. Miller, other benefits achieved include increased scalability, increased adaptability, higher accuracy, improved security and minimal false positives.
ARC research indicates that other device suppliers are developing similar approaches.
Where To Start?
Clearly field device digitization, cloud computing and new IIoT-enabled approaches offer opportunities for process industry end users to take advantage of supplier expertise in new and exciting ways. However, several roadblocks hinder widespread adoption.
Many end users hoping to reap the benefits of digitization are asking themselves: “Where do we start and how do we roll-out successful pilot projects?” It is important for users to start small with a proof-of-concept project and not try to “end world hunger” overnight. Before embarking on a large-scale roll out of digital valve management solutions, end users should develop a step-by-step plan, along with measurable goals for each step in the process.