I/O Systems

Programmable I/O can be a multitasking master

Software-configurable I/O saves on cabling, cabinets, hardware, labor and maintenance, but that's just the beginning.

By Jim Montague

One good thing leads to another, and nowhere is this more true than with software-configurable I/O. Though it can go by several different names, this quickly emerging and maturing I/O technology is granting previously unheard of flexibility when bringing process signals, data and wires into I/O points, cards, modules and cabinets. This flexibility is allowing developers, suppliers, integrators and users to simplify and save on cable and cabinets, program I/O remotely, and achieve new efficiency, optimization, maintenance and performance gains.

However, not only does assigning I/O roles using software save on hardware, it also enables users to consolidate and eliminate much traditional infrastructure; take design, planning and implementation tasks off costly critical paths for their users; virtualize process calculations and data analysis on rack-mounted or cloud-based servers; and even participate more fully in the Industrial Internet of Things (IIoT).

"Basically, a programmable I/O card makes any channel do whatever it wants, and align its I/O configuration with wires coming in from the field," says Thad Frost, technology director for I/O products at Schneider Electric, which makes FBM 247 Intelligent Marshalling modules. "Many times, customers specify having a particular percentage of spare I/O channels for each signal type. Well, a software-configurable I/O card is like having a wild card that can be any type of signal. This means less capital costs, better inventory management because fewer different backups are needed, and smaller footprints than required by dedicated I/O cards."

Simplify and save

To help update its own project specification and implementation process, engineering and construction provider Fluor Corp. in Irving, Texas, recently began using smaller I/O junction boxes with non-incendive wires, wireless instrumentation and networking, smaller remote instrument enclosures, transmitters bought and configured in bulk, and fiber-optic cabling to minimize cross wiring and marshalling.

“Labor is expensive in southern California, so it can make sense to spend a little more on materials,” says Adrian Lee, control systems director at Fluor. “Remote I/O radically reduces wiring connections and the costs that go with them."

For example, Lee explains that wiring 12,500 I/O in the conventional way—with 350,000 terminations and 201,000 wire markers at a rate of 15 minutes each—would take 1.4 million man-hours. Using Electronic Marshalling with CHARMs from Emerson Process Management, would reduce terminations to 65,000 and markers to 27,000, which requires about one-sixth the time. This saves money, but it can also shorten entire project schedules.

Likewise, it also pays to optimize the size and number of junction boxes. Using more but smaller boxes, placing them closer to instruments, and connecting them with fiber minimizes wiring run lengths. For those 12,500 I/O, using 48-CHARM boxes instead of 96-CHARM boxes adds about $785,000, or about 10% to the DeltaV DCS's hardware costs, but reducing average wire runs to 100 ft from 150 ft saves more than $16 million in wiring labor and materials. Also, using non-incendive wire means it requires no conduit, no barriers and no permits for field work, so it has the lowest overall installed cost. As a result, where common Div. 2 wiring would cost $114 million, non-incendive would cost $42 million. Fluor also found it could eliminate loop drawings by letting the connection drawings become the loop drawings, and simplifying the connection drawings by using CHARMs for a net savings of 22,000 engineering hours at $100 per hour for a total savings of $2.2 million.

“Smaller junction boxes with non-incendive wiring are the clear winner for this project,” adds Lee.

Rune Reppenhagen, DeltaV hardware product marketing manager for Process Systems and Solutions at Emerson, reports that, "Electronic Marshalling has been well adopted with more than 1,000 field installations and more than 10 billion hours of operating experience. One of its major benefits is removing the need for traditional marshalling and associated cabinets by offering a wide selection of different I/O types and field devices to eliminate the need for external signal conditioning. Electronic Marshalling provides flexibility by allowing field enclosures for the I/O to be designed before the accurate I/O mix is known. It also reduces project risk by allowing non-intrusive, flexible changes and additions of I/O without causing major redesigns and change orders. Also, by supporting nearly every possible I/O type customers would need, as opposed to only the four traditional I/O types that some programmable I/O solutions are limited to, there's little risk of needing to add external signal conditioning to a design, which would require added cabinets and enclosures."

Want More? Download our I/O Systems & Networking State of Technology Report

In the past few years, Electronic Marshalling and CHARMs evolved from their initial I/O types launched in 2009 to include intrinsically safe (IS) versions and I/O subsets, a safety instrumented system (SIS) platform and an IS version for SIS, and these CHARMs can be easily installed in junction boxes in the field. "If you're looking at a greenfield project, the number and types of I/O can vary among analog input (AI), analog output (AO), digital input (DI) and digital output (DO), but 60-80% will still be one of these four main types," says Reppenhagen. "So our work has focused on removing the need for traditional marshalling by developing and incorporating new CHARM types and terminal blocks for power, IS and other specific I/O requirements to further reduce the need for external signal conditioning. Now, we're complementing our initial I/O types with better distribution in the field by offering CHARMs in smart junction boxes that are closer to the devices they're serving, so users can handle late changes there, too."       

{pb}

Naming is the game

Just like adding other ingredients and flavors to vanilla ice cream after bringing it home, software-configurable I/O's initial, primary talent is that its wires and modules can be standardized and installed before their specific designations and roles are decided and assigned later. This seemingly small change opens new realms of flexibility, and paves the way for all of software-configurable I/O's other benefits and gains. For instance, Universal Process I/O from Honeywell Process Solutions consists of software-based programming and engineering tools that let users go to each module's 32 channels and decide which is AI, AO, DI or DO. Honeywell also plans to add a pulse-input DI to Universal Process I/O by the end of 2016, which will allow it to perform sequence-of-events functions, find first alarm triggers, and help determine why certain events happened, similar to root-cause analysis.

"Universal I/O's standalone benefits are major because they can reduce project deployment costs. Usually, there are a lot of different I/O point types that need a separate card for each role," says Joe Bastone, marketing manager for Experion controls and I/O at Honeywell. "Now, with Universal I/O, we just build one module type into process cabinets. It has 64 redundant I/O points that can be deployed anywhere, and then all their data can be brought back by fiber or IO-Link. This means less need to worry about junction boxes, homerun cables or marshalling cabinets, or the labor and cost of wiring them up." (Figure 1)

Bastone reports that a Honeywell cost study found that wiring a marshalling cabinet with I/O talking to field devices using Universal I/O with junction boxes would cost about 30% less to install than one using traditional, dedicated I/O and wiring. The study also confirmed that Universal I/O also made it much easier to recover from late changes or errors.

"For example, if a channel was misidentified, Universal I/O just reassigns it onscreen and reloads it. A traditional channel would have to be physically rewired," says Bastone. "It helps potential users to see this difference in practice, so we demonstrate it at our development centers, and many customers order a couple of cabinets to prove it to themselves. We're not really inventing new processes, hardware tools or form factors; we're taking existing principles, such as all our Series C products, and extending what we've already done with Universal I/O."

Schneider's Frost adds, "The biggest gain for programmable I/O is that it allows users to break their dependence on traditional control system design and I/O system installation. Typical systems engineering requires defining and designing in all the different I/O points they need, but programmable I/O can be designed at the same time as I/O installation onsite.

"Likewise, about 90% of process control projects are installed around unknown elements, so users don't know exactly what I/O they'll need. However, programmable I/O let's them put in what they can now and decide on transmitter, level switch or other roles later, or even change them on the fly when their application is running without taking down the whole I/O card. Traditional I/O loading can be a very mundane and thankless task, but with programmable I/O, much of it goes away."

On beyond marshalling

While software-configurable I/O makes landing cables, marshalling and cabinet building simpler and easier, its flexibility has been helpful in other process applications as well.

To coordinate refining processes on its Mobile Alkane Gas Separator (MAGS) for oil wells, Pioneer Energy in Lakewood, Colo., uses programmable I/O modules and SNAP PAC controllers from Opto 22. A truck-mounted MAGS captures raw/wet natural gas that would otherwise be flared off at recently developed oil wells, and processes it into usable lean methane, ethane and Y-grade natural-gas liquids (NGLs), such as propane and butane (Figure 2). Lean methane is 80-90% methane and can be used directly in generators and other natural gas-powered engines, while Y-grade NGLs have a low vapor pressure of no more than 250 psia at 100 °F, and can be transported in regular propane trucks.

The compact compression, dehydration, refrigeration, three-stream separation and other operations on MAGS must work together with low power requirements, and support a range of I/O signals. Opto 22's components automate, monitor and control MAGS' temperature sensors, flowmeters, pressure transmitters, control valves, generator, refrigeration compressors, gas chromatograph, process heaters and other equipment.   

“To interface with all the equipment on MAGS, we use nearly all of Opto 22’s I/O modules and SNAP PAC's built-in Modbus protocol support,” says Andrew Young, lead controls engineer at Pioneer. “Pioneer focuses its technology on adding value to waste streams. We see flare gas as another potential waste stream, and remote communications and a reliable controls platform let us convert that into a resource.”

Because many well sites are remote and unmanned, Young reports that MAGS’ programmable I/O and controls must run reliably and unattended, and support multiple communications protocols to operate autonomously. To commission or troubleshoot MAGS in the field, Pioneer also uses Opto 22 groov mobile operator interface system, which lets users connect securely to MAGS’ control system with a tablet or other mobile device, and adjust settings in real time. Pioneer's Lakewood operations center also uses Opto 22 PAC Display HMI software to remotely monitor and manage MAGS units in North America.

"Despite a huge installed base of legacy PLC and relay-based control systems that use non-programmable slave I/O, increasing numbers of engineers have evaluated and adopted intelligent, programmable I/O systems because they’re easier to use, architecturally versatile and network-efficient, and help process control systems become safer and more reliable," says Matt Newton, technical marketing director at Opto 22. "Higher performance in programmable I/O has been achieved primarily through two main areas: networking and faster CPUs. More advanced and higher-speed networking options are continually being developed. Redundant ring architectures and gigabit Ethernet speeds are two examples. Faster processors are also continually developed and deployed that improve the speed of logic within programmable I/O. These advances continue to improve the overall performance of programmable I/O."

Randy Durick, vice president of fieldbus technology at Turck USA, explains that signals and data previously passed along via fieldbuses and Ethernet to centrally located logic solvers are often giving way to local network gateways and couplers that can do more computing on their own, and then report up to central controls as needed.

"Formerly, Foundation fieldbus did control in the field by using function blocks to control final elements, but more recent migrations are using distributed I/O and electronic marshalling like CHARMs that are inside and outside of cabinets," says Durick. "These gateways and couplers have embedded programming like IEC 61131-3 with logic for analog and discrete tasks, as well as runtime kernels that are standardized across manufacturers. We're seeing these components in oil and gas—not so much in complex downstream applications, but more in simpler upstream applications with fewer I/O, such as our BL20 gateways backing up 20-40 I/O points that are simpler to assign and easier to deploy."

In fact, Durick adds that Turck has just released its new Field Logic Controllers, which can be programmed via a web browser and support HTML 5. This is how it accesses Turck's ARGEE engineering environment, which can then program the controller's embedded firmware. 

{pb}

Power conversion and other snags

Though the potential benefits of software-configurable I/O are enormous, there can still be some power conversion, networking and other technical hurdles to overcome on the way to securing those gains.

"Not having to use traditional, dedicated, 16-channel I/O cards allows users to configure 32 available I/O points on the fly, make late design changes, customize those points by function, and make them what they want," says Jeremy Valentine, product marketing manager for interface components at Phoenix Contact. "This can have a huge impact on project planning cycles. Traditionally, if you needed 5,000 I/O points, plus or minus 10%, then you'd have to do the math early on the required numbers of analog or digital points, and there was a heavy burden on precision and signing up the right dedicated equipment. And, if you had to change the amount of customized, dedicated devices later, it could be a big problem. Universally configurable I/O eliminates this, and shortens project design, execution and lifecycles by allowing programming changes to be made later. Enabling these late-binding changes can avoid a lot of 'oh crap' moments, and shorten project lifecycles from 18 months to six or eight months."

Valentine adds that Phoenix Contact and other suppliers are seeking ways to make software-configurable I/O platforms even more universal by solving some of the first-generation issues many face, such as a frequent inability to handle 120 V or more than 3 A power, being required to convert to 24 V, and needing relays to handle these low-current-capacity limits. "At the circuit board level, everyone wants to add more components, but that requires more and better surge protection, carrying more current, and 4-20 mA signals coming to analog cards though conversion isn't available," explains Valentine. "Unfortunately, the limit is still 3 A on the on the digital relay side, so if a higher load needs to be driven, then another contact will be needed. As a result, though suppliers say they can eliminate marshalling and cabinets, issues like the 120 V-to-24 V limit, means some of the required technology isn't quite there yet. Still, in the next five or 10 years, the conversion abilities will likely be on the boards as well."

Albert Rooyakkers, CTO and engineering vice president at Bedrock Automation, notes that users also need to question what their programmable analog I/O devices can support at the levels where they're applied.

"What resolution, accuracy and isolation do they need? Do they require a fast channel response? HART? What discrete I/O are they using, do they support NAMUR, and how much current can they handle?" asks Rooyakkers. "Do they need power to source a discrete input? Are they able to do pulse? Universal I/O should start with this level of capability to be truly valuable to users."

Rooyakkers reports that all 10 channels of Bedrock's SIOU.10 universal I/O module are galvanically isolated and software-programmable to receive field input or output signals for analog, discrete or pulse sensors and actuators. He adds that Bedrock's SIOU.10 also has integral read-back functions to confirm analog and discrete outputs. “Virtual marshalling via universal I/O is the new standard for control system engineering and optimal life cycle ownership," he says. "With deeply embedded cyber security, scalable redundancy and unprecedented technical specifications, SIOU.10 combines the equivalent functionality of multiple I/O module types into one software-programmable module."

Off the critical path—and upwards

No doubt the greatest benefit of software-configurable I/O is that it can take design, planning and installation projects off the costly and risky critical path they're usually on during implementation.   

"When big jobs are executed, concrete is poured, and process equipment is in place, then the last thing on the schedule is the DCS," says Schneider's Frost. "However, dedicated and inflexible wiring, marshalling and cabinets could make this risky, especially when late changes were needed. The flexibility of programmable I/O allows many jobs to be done in parallel instead of serially, takes them off the critical path, and help avoid calls we want to avoid, such as, 'Why isn't my plant running?'"

Beyond saving hardware and time, programmable I/O allows the sensors and instruments it serves—and the process-level signals and data they deliver—to cooperate more closely with higher-level networks, controls, analytical and business systems.

"The second path for Universal Process I/O is that combining it with virtualization and cloud-based engineering systems as part of Honeywell's LEAP program means that user can rethink their designs, engineering, implementation and deployment," says Bastone. "Universal I/O also allows users to extend their hardware and software freeze dates, which is when project configurations must be frozen in time before construction and installation. Universal I/O and LEAP also allow simpler, standard engineering drawings, which lets users consolidate on several cabinet types and create standard loop templates. This is better than creating 20,000 loop drawings for 20,000 I/O points, and building custom cabinets and maintaining them throughout the lifecycle of a project. If you can reference a master list of 20 loop templates, then you'll know exactly how to wire up each one."

Similarly, Emerson's CHARMs fit into its overall Project Certainty program, which begins with early engagement during clients' engineering and design studies to define project goals and high impact strategies to meet those goals. Kevin Jackson, vice president of global project operations at Emerson, reports that Project Certainty eliminates labor costs by eliminating unnecessary work, improving the efficiency of necessary work and eliminating rework, and it cuts material costs by eliminating the need for piping, structural and electrical components, and by pushing for "fit for purpose" and right-sizing engineering on projects.

"On a recent project that previously had 18,000 I/O and needed 32 cabinets, Electronic Marshalling with CHARMS reduced traditional I/O by 37% to 20 cabinets, and Electronic Marshalling with Smart Junction Boxes reduced this project another 95% to just one cabinet for a total savings of $14.5 million," says Jackson. "Likewise, when a safety measurement requirement for pump control and shutdown was added late in a project design cycle, an EPC saved about $1 million thanks to a 97% reduction in straight-run piping on four LNG trains (Figure 3).

"In addition, a recent, integrated well-test solution saved $3.5 million on seven units by eliminating separators, valves and instrumentation; reducing piping by 60%; and reducing engineering, construction and leak points. Another end user eliminated $23 million in capital expenditures for spare parts by identifying common and critical parts, and performing a cost/risk analysis on them."

Project Certainty works in conjunction with two other new Emerson programs: Project Data Link, which is a single-source-of-truth engineering environment that helps reduce complexity and accommodate changes in capital projects, and Smart Commissioning, which builds on the combination of the Electronic Marshalling with CHARMs and AMS Device Manager software to remove automation from the critical path. "Smart Commissioning reduces typical loop commissioning times from 140 minutes before to 25 minutes now to 10 minutes in the future," says Jackson.

Frost adds that Schneider Electric's umbrella initiative for compressing or even eliminating factory acceptance tests (FAT) and other pre-startup tasks is called Flexible Lean Execution (FLEX) program, which includes Intelligent Marshalling, universal I/O, engineering in cloud-based services and virtualized computing. These capabilities are also based on Schneider's Foxboro Evo process automation systems and Triconex process safety system.

"This is why there's been such an uptick in interest in having I/O undefined at installation, and then configured in software," adds Frost. "Programmable I/O can slash capital costs, which also helps when oil prices are down."

Cloud, virtualization and IoT coming

Naturally, the ultimate expression of software-configurable I/O's flexibility and simplicity is how they help process control applications run in virtualized, cloud-based and IIoT systems.

"As IIoT grows, programmable I/O will play a large role in its architecture," explains Opto 22's Newton. "Key capabilities for a workable IIoT solution are translation of physical things or properties into digital data, edge processing, protocol translation, and fast, dependable, ubiquitous networking and communications options. Programmable I/O possesses all of these in a compact, industrial package. Future programmable I/O will add native IIoT communication capabilities such as RESTful interfaces, IIoT protocols like MQTT, data storage and analysis, visualization of data, cloud communications and more."  

Emerson's Reppenhagen adds, "As we evolve this technology further, I think we're going to see more flexibility for connections between I/O 'clouds' and controller 'clouds.' There's more and more use of Ethernet in automation today; for example, last fall we introduced an Ethernet I/O card to enable large data transfers using Ethernet, and our partner Mynah Technologies also recently released a driver for our Virtual I/O Module 2 that now supports Profinet natively in DeltaV. It will be interesting to see if there will be a merging of Ethernet protocols in the future, or if it will be similar to the bus technologies of the past with multiple protocols on the market for the long term. WirelessHART is also a well-proven technology that was first introduced for monitoring only, but is now used in some applications for control. Wireless I/O is the ultimate programmable I/O in that it allows instruments to connect to wireless I/O cards without wiring and with no marshalling, and it just works for all wireless I/O types."