Special Section: Wireless & Ethernet
Ethernet empowers fieldbus
Fieldbuses are adapting Ethernet to increase performance, cut costs
By Craig McIntyre
Fieldbus networks have been around for over a decade, delivering value in industrial automation applications worldwide. In most cases, a fieldbus network is used to link field devices to a host computing system in a process plant.
The field devices in question are most typically instruments, analyzers, and modulating control valves. The most popular types of instruments are flow, level, temperature, and pressure transmitters—although a typical process plant will also include many other types of process variable transmitters.
Common analyzer types measure the chemical composition of parameters such as moisture, carbon dioxide, methane, and other gasses and liquids. Modulating control valves provide continuous flow control of liquids and gases.
The host computing systems linked to field devices include control systems, Enterprise Resource Planning (ERP) systems, and asset management systems. The control systems in question are usually basic process control or regulatory control systems. Other control systems with fieldbus links to field devices include advanced process control systems and safety systems.
The benefits of linking field devices to host computing systems via a high speed data link like fieldbus include:
- Networked device configuration and health management saves money
- Networked device documentation saves money
- Predictive maintenance increases uptime
- Predictive maintenance improves performance
- Predictive maintenance cuts maintenance costs
In almost all cases, fieldbus is used to replace a hard-wired 4-20mA one-way connection from the field device to the host computing system. This 4-20mA signal was used to transmit the measured process variable to the host computing system.
In contrast, fieldbus provides a high-speed two-way data link that can transmit copious amounts of information between the field device and the host computing system. The most important bits of information are data related to instrument health, which can be used by the host computing system to schedule calibration as needed and for predictive maintenance.
Scheduling calibration only as needed instead of on a periodic basis saves money because field devices are not calibrated when operating within parameters. As needed, calibration improves performance because more frequent calibration can be performed on critical field devices that are drifting out of range.
Predictive maintenance is perhaps the most important benefit delivered by a fieldbus system. Using the data delivered by fieldbus, a plant can predict problems before they occur. Maintenance can then be performed on a planned basis as opposed to a reactive basis, saving money and improving safety.
But the main benefit of predictive maintenance is avoidance of downtime. If a field device is found to be failing, it is often possible to repair or replace the device before it brings down the entire process.
Fieldbus is a generic term for a number of digital industrial networks, including but not limited to, Foundation Fieldbus, HART, EtherNet/IP, Modbus TCP, and Profibus. Many fieldbus networks, including those mentioned above, are transitioning to Ethernet-based protocols for a variety of reasons.
Benefits of fieldbus have been well documented in many process plant applications for years, and most of these fieldbus installations use proprietary protocols as opposed to Ethernet-based protocols. But Ethernet-based fieldbuses can often provide better performance at a lower price, while also simplifying installation and maintenance.
When the term Ethernet is mentioned in the commercial world, it is usually in reference to an Ethernet network utilizing the TCP/IP protocol. In the industrial world, the term Ethernet only identifies the underlying hardware and not the protocol, which can be found in various flavors. The bad news is there are many competing industrial Ethernet-based protocols. The good news is they can all run on the same underlying Ethernet hardware, often simultaneously. (See sidebar for more information.)
An improved price/performance ratio is perhaps the main benefit of switching to Ethernet. A few years back, one of the leading fieldbus network organizations was trumpeting the fact that over 1 million fieldbus devices using their proprietary protocol had been sold over its 10-year life. This number may have been impressive compared to other proprietary competitors, but it is trivial in comparison to the number of Ethernet devices installed worldwide.
As the leading protocol for computing connectivity, there are billions of Ethernet nodes installed worldwide, mostly in commercial applications. Large numbers generate economies of scale, allowing providers of Ethernet hardware to continually drive down costs and increase performance. And because Ethernet is a worldwide standard, vendors compete based on price/performance ratios and thus have tremendous incentive to deliver the best bang for the buck.
Fieldbus organizations have taken advantage of commercial Ethernet economies of scale to deliver higher speeds at lower costs. Higher speeds allow quicker update times for monitoring applications. Depending on the characteristics of the process being controlled, higher speeds can also enable real-time control.
Because Ethernet is so pervasive, installation and maintenance are easier for a number of reasons. First, it is much easier to find technical personnel familiar with Ethernet as opposed to a proprietary network. Second, hardware and software tools for installation and troubleshooting are widely available at low cost. Finally, it is often possible to lean on corporate and plant IT personnel for support and maintenance because IT folks speak Ethernet. By converging their industrial and office networks, end users have fewer variant networks to maintain and gain more leverage when integrating technologies and communications.
Ethernet provides concrete benefits to fieldbus, and many users are taking advantage in a variety of applications. In particular, Ethernet connectivity is bringing fieldbus to the Programmable Automation Controller (PAC) level.
Fieldbus Ethernet enables PACs
In early implementations, most fieldbus control systems were of the DCS variety. Now that Ethernet is providing a common communications protocol, fieldbus is becoming a viable option for processes controlled by PACs. This can decrease costs for process control systems as a PAC is typically less expensive than a DCS.
A leading manufacturer of floor coverings for commercial and residential applications uses a combination of PACs and EtherNet/IP-enabled field devices to improve product quality, promote green manufacturing methods, and enhance production efficiencies. At one of their plants, this manufacturer uses a PAC as their main real-time control platform. The PAC comes with EtherNet/IP built in, so the plant would like to use this protocol as their Ethernet-enabled fieldbus.
Field devices like these flowmeters deliver more value when connected to control systems, asset management systems, and ERP systems via an Ethernet-enabled fieldbus.
Consequently, an EtherNet/IP-enabled Coriolis flowmeter was required for measurement and control of mass flow in one of the plant’s continuous processes related to colorant control. The mass flowmeter selected was found to have superior accuracy in competitive field trials conducted by the manufacturer. Direct connectivity from the meter to the PAC via EtherNet/IP provides a number of benefits.
The meter is capable of simultaneously measuring multiple parameters including mass flow, product density, process temperature, volume flow, custom concentration, and viscosity. The plant wanted the ability to monitor these parameters without having to run multiple wires, and the solution was the high-speed 100 Mbps EtherNet/IP protocol.
The meter’s advanced diagnostics parameter monitoring can now be used by the plant to predict process influences from coating, buildup of solids, corrosion, erosion, and entrained gas conditions. Predicting problems before they occur enables predictive maintenance, cutting costs, and reducing downtime.
Further benefits identified in this application included a 40% reduction in device commissioning time and a 25% reduction in loop identification, device integration, and process loop tuning time. Immediate recognition of the meter as a network node is another benefit, along with transparency of the meter from the factory floor to the enterprise system.
There is no doubt that fieldbuses will continue to evolve towards Ethernet-based implementations. To remain competitive, the fieldbus organizations will have to take advantage of Ethernet’s superior price/performance ratio.
End users will demand Ethernet-enabled fieldbuses to simplify direct connections from field devices to host computing platforms such as ERP and asset management systems. For the growing number of process plants using PACs, Ethernet-enabled field buses will be seen as a natural fit.
More and more plant floor technical personnel will become familiar with Ethernet, encouraging its use in process plants. Internal IT personnel will also prefer Ethernet-enabled fieldbuses because of similarities with corporate computing system networks. The Ethernet-enabled fieldbus bandwagon is rolling forward, and suppliers and end users are jumping on.
ABOUT THE AUTHOR
Craig McIntyre (firstname.lastname@example.org) is the chemical industry manager with Endress+Hauser in Greenwood, Ind. Other positions he has held with Endress+Hauser during the last 17 years include level product manager, communications product manager and business development manager.
Direct is best
When flying from one city to another, direct flights are always better than those requiring an interposing connection at a hub airport. It is much the same with fieldbus connections to higher level host computing systems, where direct connection from the field device to an ERP or asset management system is preferred to a connection via an interposing control system.
Bypassing the control system and going directly to a host computing system like an ERP or asset management system is made easier by Ethernet-enabled fieldbus. That is because most every host computing system is capable of Ethernet communications.
In a typical process plant, most field devices are not directly associated with real-time control and are instead used primarily for monitoring. Some facilities have found up to 70% of their field devices do not have any associated control functions. It is no longer correct to assume every field device must be connected to a control system to have value.
For example, environmental applications such as EPA monitoring and reporting require multiple field devices that do not need to be connected to the control system. ERP applications such as inventory management need field device input, but do not require the millisecond update speeds or deterministic behavior associated with control systems.
Bypassing the control system provides a number of benefits. First, a direct connection eliminates the need for intermediate hardware components and software systems. Second, the field device in question may not need to be part of the control system’s overall validation and maintenance program. Third, field device access can be controlled through existing IT security systems.
Information from these non-critical field devices can be conveyed directly to process monitoring applications via standard Ethernet networks and wireless field gateways. Not only is the primary information delivered in fully defined engineering units, but device status can be continually monitored and communicated on an event driven or periodic basis.
Remote servicing tools and asset management applications working at the network level can also configure and manage connected devices. Standard IT security and data management tools can be used to control access.
All fieldbus Ethernets are not created equal
Fieldbus organizations like to boast that their particular flavor of Ethernet network is the best, adhering most closely to commercial Ethernet TCP/IP implementations while still delivering the real-time performance and reliability needed for industrial applications. While those claims are for end users and system integrators to judge, there is no doubt Ethernet-based fieldbuses come in many different and often incompatible varieties.
For example, EtherNet/IP and PROFINET are Ethernet-based fieldbuses, but it is not possible to mix and match components adhering to these two standards without some kind of gateway and/or translator. In other words, a PROFINET-enabled field device cannot be directly connected to a Programmable Automation Controller (PAC) with an EtherNet/IP port.
One solution to this problem is to only use field devices and controllers compatible with one fieldbus. Unfortunately, this is often not a viable option as most process plants have existing field devices adhering to different fieldbus standards.
Even for a greenfield plant or process, it is usually not possible to purchase all of the required field devices from vendors adhering to one Ethernet-enabled fieldbus standard. Certain specialized instruments, analyzers, and control valves are often needed, and these field devices need to be connected to the control system and fieldbus of choice.
To cope with this issue, many vendors make gateway devices that convert one Ethernet-based fieldbus protocol to another. For example, a gateway can allow connection from a Modbus TCP-enabled field device to a PAC with EtherNet/IP connectivity.
Although gateways solve the incompatibility issue, it is best to minimize use for a number of reasons. First, gateways add to the overall cost of the fieldbus installation. Second, gateways add complexity in design and maintenance. Finally, gateways require additions to maintenance inventory and increase stocking requirements.
There was hope Foundation Fieldbus High Speed Ethernet would become the industry standard, but not all DCS vendors subscribed to this effort. End users long for the day when one Ethernet-enabled fieldbus emerges to rule them all, but until that time, it is best to use compatible components to the greatest extent possible, with gateways accommodating outliers.