Remote and Autonomous Operation - No Routine Visits

Imagine, unmanned offshore platforms and other remote locations, with no routine visits. This is remote and autonomous operation. For plants by urban areas, remote and autonomous operation may mean nightshift and weekend operation with skeleton crew with no routine visits to the field. Since the first Friday in October is the #ManufacturingDay I decided to look at how digital transformation including IIoT can help production. The production process is already very automated, just needs to be remoted. Maintenance and reliability need much more attention. How do you transform plants from local operation to remote and autonomous operation, and how long can we go between visits? Here are my personal thoughts:

Lights-Out?

‘No’ and ‘Zero’ goals are typically aspirational. No plant can be completely autonomous. It may be unattended for long periods of time, but eventually somebody will have to go there. So remote and autonomous operation (R+AO) is about extending the period between site visits for as long as reasonably possible.

Simple processes can run unattended for long periods without human intervention such as oil & gas production, water and wastewater treatment, and industrial gas production (air separation). It may need some human supervision from time to time but it can be from a central location managing multiple sites. More complex process units require human attention such as in a refinery, so production operators are always present. This may also include an offshore mother platform serving multiple satellite platforms or a Floating Storage and Offloading (FSO), Floating Production Storage and Offloading (FPSO), Floating Drilling and Production Storage and Offloading (FDPSO), Floating Liquefied Natural Gas (FLNG) production, Floating Storage Regasification Unit (FSRU), and even floating power plants. However, even at sites with complex processes, field operator routine visits can be reduced by doing surveillance remotely from a central location. That is, the site has a smaller crew, and a larger portion of personnel working onshore, and fewer people offshore.

Reliability tasks like inspection with portable testers and reading gauges can be fully automated and thus autonomous. But once a problem is predicted, maintenance needs to send a human there for the physical repair. Same for replenishing consumables.

Operational Excellence

There are many reasons sites are looking to transform their remote operations. Many plants are high risk workplaces, particularly offshore. There is a need to keep people out of harm’s way. It is hard to attract and retain talent to remote isolated sites, particularly in the desert or arctic, or high-risk regions. The cost of accommodating personnel offshore or in camps at remote locations is very high. Transporting personnel as well as vendors and contractors by plane, helicopter, or boat for routine or ad-hoc tasks is also very expensive. The cost of frequently transporting personnel and other logistics is high. Travel is not productive. Onshore labor rates are lower than offshore labor rates. Therefore the need for reducing the number of people living and working at site by moving activities offsite and enabling new ways of working. Moreover, getting an expert on site can take long causing delays. COVID-19 demonstrated epidemics and pandemics complicate matters. For a new offshore installation there is significant savings in not having permanent accommodation.

Remoting and Automating Work

The process industries are tackling these challenges by relocating staff from high-risk locations like offshore installations to low-risk locations like an urban office, and by reducing the need for manual intervention in the first place. Part of this vision is to also enable personnel to work from home. There are two parts to operation: production and maintenance. Together they are called integrated operations; iOps for short.

Unmanned offshore satellite platforms have long been operated from mother platforms so the concept of reduced manning or unmanned is not entirely new. However, it has been limited to running the production. We can call this “fleet management 1.0”. Other functions such as maintenance, reliability, and integrity still required people to be transported offshore. In the past remoting these functions was not practical with the technology at the time, but now the enabling technology is available to centralize the maintenance, reliability, and other management of plants across the world. We can all this “fleet management 4.0”. Automation is now available for autonomous unmanned operation for sustained periods of time, or reduced staffing. Thanks to new automation plants can run for as long periods as possible with no travel to site. Vendors and contractors provide support without going to site. Many of the activities traditionally performed locally at site are instead being done from a fleet management center. Moreover, periodic inspection instead becomes continuous monitoring making operations more predictive. Digitization enables operations to put new work processes in place. Instead of roaming the site, personnel work from a fleet management center.

How long between visits to the site depends on the type of plant, and the initial condition of the plant.

Production, Safety, and Quality

Production at multiple unmanned sites is centrally supervised from a fleet management center. In an existing plant the process is already controlled by a DCS, PLC, or RTU. Process safety by a Safety Instrumented System (SIS). Core Process Control (CPC) and process safety (SIF) shall remain on site for best performance and reliability. There may or may not be very much to be done w.r.t. process control. It’s mostly about extending the control system operator console to the fleet management center. This can be achieved using the operating system remote desktop functionality which enables full DCS supervision including setpoint, controller mode, manual output changes, open/close, and start/stop. This way the console operator can work from the fleet management center instead.

The SIS must be autonomous, able to function when the communication link between plant site and fleet management center is lost. Operators in the fleet management center need the ability to force shutdown at the production site. Additional sensors for liquid and gas detection can be deployed.

An autonomous SIS also requires automatic testing of the SIF loop end-to-end. For some plants the field instruments must be upgraded to smarter versions with self-diagnostics. For instance, transmitters detecting sensor failure, a Partial Stroke Testing (PST) device may need to be fitted on the shutdown valves, and there must be current loop diagnostics for 4-20 mA loops. This instrument diagnostics must be communicated to Intelligent Device Management (IDM) software. The SIS may have native integration with the IDM software or may integrate using HART-IP. Older SIS has no support for HART and require a multiplexer (MUX). If a MUX is used, it should be based on HART-IP. Additional HART filter may be required for the PST device.

Plants should be automated to a very high degree to be autonomous. Shift, daily, weekly, and monthly field operator surveillance rounds should be eliminated for unmanned sites. There already are unmanned sites today where a field operator is not present most of the time, but today most unmanned sites must be visited too frequently, maybe weekly. There is a need to extend the period between visits by automating the data collection. Even for normally manned plants the shift and daily field operator surveillance rounds should be eliminated to enable minimal crew during nightshift and weekends. That is, during the nightshift and weekend there are no routine visits out into the field; nobody leaves the control room.

Grab sampling to bring product sample to the lab must also be eliminated for the site to be autonomous. If available an inline analyzer to measure product composition and properties directly might be the best option. If not, consider inferential measurements based on direct measurements by installing add-on sensors for pressure, temperature, density, pH, and conductivity.

In the future, video cameras, thermographic imaging cameras, and LiDAR based on the future Ethernet-APL can be dropped in to monitor the process, flare, or detect leaks.

Maintenance, Reliability, and Integrity

Maintenance and reliability today are very manual. What today is referred to as unmanned sites require frequent visits for routine inspection and servicing. Additional automation must be deployed at site to make maintenance and reliability autonomous and enable it to be managed and planned from a central location and get to no routine visits to site. A condition-based maintenance (CBM), predictive maintenance (PdM), or prescriptive maintenance (RxM) regime is required to reduce routine and repair visits to site. Without condition monitoring, an unmanned site will experience many unplanned shutdowns. Maintenance such as lubrication, alignment, cleaning, and replacements and logistics such as replenishing consumable are only performed when predicted or detected necessary. That is, extending the period between visits by moving from routine visits to condition based. With centralized maintenance management in place, manned sites can have a smaller standby maintenance crew during the nightshift and weekends. The health of equipment in multiple plants are monitored from a fleet management center so when they predict a developing problem with an equipment they can decide to shutdown the unit or instruct an operator to fix it, possibly guided step by step by an expert in the fleet management center.

Once digital, all or part of reliability and maintenance management can optionally be outsourced as a connected service. But at the same time, reliability and maintenance surveys that used to be outsourced to third-party service providers which had to visit the site can now be done without coming to site.

Discovery Workshop and Solutions Mapping

In order to digitally transform the site to enable remote and autonomous operation you need an understanding of why personnel and contractors have to be transported to site or why they have to go into the plant, especially at night or over the weekends. Therefore start the transformation journey by conducting a discovery workshop to review these site visit tasks part of the duties of the roles in each department at a site. What routine manual inspection and surveillance rounds are there? What are the ad-hoc callouts? To check if something is functioning properly? Or do they put an operator on standby in the field monitoring some activity? Is a single or multiple trips required? Maintenance, reliability, integrity, process energy, HS&E, production, and quality departments all have their reasons to go. These reasons and frequency need to be captured. Sites have records of personnel transport that can be reviewed to assist in this discovery. Similarly, most plants have records of people going to the field. Depending on how long you target between scheduled site visits the workshop would need to review the tasks performed daily, weekly, monthly, quarterly, and yearly to uncover what would need to be automated or remoted.

Based on the discovery workshop results it is possible to determine what can be automated and what can be supervised and managed from the fleet management center and map that to readymade solutions, such as monitoring and analytics, or create new solutions as needed. In case unexpected site visit is required, the solutions should make sure they don’t have to comeback a second time; it should be a first-time fix.

The discovery workshop also captures when manual intervention by console operators is required. All control loops should be in automatic mode as it is not practical to have many loops in manual. Why are loops in manual? Is it due to poor loop tuning? Maybe poor valve performance? Surprise process upsets?

Based on the discovery workshop results it is possible to determine a solution as simple as loop tuning, or valve performance diagnostics, or predictive process analytics.

Digital Operational Infrastructure (DOI)

The vision of remote and autonomous operation is made possible by new and improved automation not available in the past.

The DCS, PLC/HMI, or RTU/SCADA system at site for Core Process Control (CPC) may not need much modification to enable remote operation through remote desktop protocol. Your I&C engineers and the automation vendor are very familiar with this. It is mostly a matter of providing a secure connection to the fleet management center. The automation system should be fitted with an OPC-UA server to allow process data to be used in analytics and other functions.

As per the NAMUR Open Architecture (NOA) best practice for Industrie 4.0 deployment, most of the additional automation required for remote and autonomous operation does not connect to the process control system. Instead, NOA defines an interface to a second layer of automation on the side of the process control system. This new Digital Operational Infrastructure (DOI) for Industry 4.0 includes the additional dashboards, notifications, reports, analytics, wireless networks, and additional advanced sensors required for digital transformation of work enabling remote and autonomous operation. This separation helps maintain the robustness and safety of the DCS.

Figure courtesy: NAMUR

The DOI is rolled out gradually at sites. For new plants the DOI should be included from the very beginning. A fleet management center has visibility of all sites. If functionality is deployed at site or centrally is up to company operating philosophy. A key principle though is that the robustness of the core process control is not put at risk.

Figure courtesy: NAMUR

A DOI is more than a data management platform. In the DOI technology stack the top layer is the presentation canvas which can be a web interface like Microsoft Power BI to display dashboards with real time indexes, KPIs, and other information. This information in turn comes from the underlying application framework which hosts the analytics apps which distills the information from raw data. The raw data in turn comes from the underlying data management platform. This may include the process data in the existing historian, if there is one, as well as an aggregating server, an extension for all the new data not already in the historian. Lastly, the very foundation on which everything rests is the data sources which includes existing system as well as new wireless sensors. Automation vendors can supply the whole DOI stack.

Solution Details

To learn the details of the enabling solutions, see the earlier “Keep the Plant Producing” essay found here:

https://www.dhirubhai.net/pulse/keep-plant-producing-work-from-home-jonas-berge/

The same solutions that enable work from home, enable work from anywhere, such as a fleet management center.

Process Analytics

For autonomous operation the need for manual operator intervention during normal operation must be minimized. If there are surprise process upsets because operators don’t recognize when process disturbance are coming or don’t take correct actions, consider deploying process analytics to predict and recommend action.

Wireless Sensors

Sensors automate inspection and surveillance data collection to make a site autonomous and enable remote operation. Equipment must be fully instrumented.

When there are no people at site there are no eyes and ears. Therefore more sensors must be deployed to get that same level of situational awareness for the personnel relocated to the fleet management center.

A condition-based maintenance (CBM), predictive maintenance (PdM), or prescriptive maintenance (RxM) regime requires additional sensors to directly pick up on the early signs of developing problems which enable analytics software to predict problems well in advance.

When there is no fully automatic analyzer to take the place of lab testing, inferential measurements based on simple measurements may be the solution. This requires direct sensors for pressure, temperature, density, pH, and conductivity.

That is, sensors are the very foundation upon remote and autonomous operation rests. Selecting the right Wireless Sensor Network (WSN) technology for the plant is key to reliability, supporting the required applications, and ease of use. With the wrong WSN technology you get none of that. Your I&C engineers are familiar with the requirements for industrial sensors and networking.

Virtual Reality (VR)

With infrequent visits to sites, and the same person covering multiple sites, production operators and maintenance technicians will be less familiar with the specifics of each site.

Virtual Reality (VR) is a very immersive way for operators located far from site to familiarize themselves with the plant they will operate before they start. I’m not suggesting using VR to operate the plant. Such an immersive experience is not necessary for operation, and quite possibly distracting. But VR is great for familiarization before operation starts.

Production has many manual tasks such as opening and shut in of wells, startup and shutdown of units, but also loading and offloading. Virtual Reality (VR) enables familiarization with the site and practicing such manual tasks in the safety of a classroom before even sending personnel to site. Because the environment and process are virtual, not real, students can make mistakes without consequence so they can practice over and over as much as it takes. Operator Training Simulator (OTS) software for control room operators has existed for many years. VR now enables practicing tasks done out in the field improving production and safety. Emergency evacuation escape routes can also be practiced. The virtual three-dimensional (3D) environment is very realistic mimicking the actual site with high fidelity. It is like being there. It looks and feels like the actual site, not generic. Piping, vessels, structures, process equipment like pumps and heat exchangers, and valves are rendered in high-fidelity for a realistic experience.

You need a digital twin for the simulation, but the most interesting fact is that is the only thing that needs a digital twin. All other functions such as automating data collection, predictive equipment health and performance analytics, dashboards and notifications, AR, VR, RFID, and more do not need a digital twin. That is, a digital twin is not the solution to digitalizing all tasks, just some.

Telecom Backhaul Infrastructure

Remote and autonomous operation means more data to be communicated between the site and the fleet management center. An existing site may therefore need to upgrade the telecoms link. A new site should be deployed with a high bandwidth link right from the start.

Process controllers and safety logic solvers run locally in the plant, and edge analytics is used to minimize telecom bandwidth requirements. This also means core process control (CPC) and safety does not rely on the telecoms link with the fleet management center to function.

For greater backhaul availability the telecoms link shall have redundancy. That is, two alternate means. It could be two separate optical fibers, or any two of microwave data link, satellite, mobile network, radio, or even dial-up.

Wearable tablet for live remote assistance requires good bandwidth for the video feed. The advantage is that site personnel can make repairs that previously required a vendor or contractor to be transported to site.

No Routine Visits

The Fourth Industrial Revolution (4IR) enables working in an urban office with reduced number of personnel at the production site, even unmanned, and provides easier and faster access to experts for faster problem resolution. The result is greater reliability and energy efficiency, reduced cost for operations and maintenance, plus greater safety. Schedule a meeting for the first Friday of October manufacturing day or today. Forward this essay to your operations manager now. And remember, always ask for product data sheet to make sure the product is proven, and pay close attention to software screen captures in it to see if it does what is promised without expensive customization. Well, that’s my personal opinion. If you are interested in digital transformation in the process industries click “Follow” by my photo to not miss future updates. Click “Like” if you found this useful to you and “Share” it with others if you think it would be useful to them. Save the link in case you need to refer in the future.