When Will We Have Process Field Devices with Ethernet and IP?

Author photo: Harry Forbes
By Harry Forbes

Overview

Higher level process automation networks evolve almost in lock-step with information technology (IT), adopting internet protocols and the higher speeds of 100MB and 1 GB Ethernet, which are now commonplace.  With the process field devices in the process process field devicesindustries, however, network technology remains essentially the same as it was in the 1990s or even the 1980s.  Over the long term, this lack of technical evolution at the field level has become a pain point for many end users, and the continued use of legacy field networks (HART, FOUNDATION fieldbus H1, and Profibus PA) will only increase the pain over time. End users are attracted to the vision of simplified commissioning via networked field devices that embed their own configuration data (such as the device DTM or FDI package).

At a session of the 2016 ARC Industry Forum in Orlando, speakers from BASF and Pepperl+Fuchs discussed the future of Ethernet and IP networks in process field applications. The session speakers were Dr. Michael Krauss, Senior Automation Manager at BASF, and Dr. Gunther Kegel, CEO of P+F.   The two took turns discussing their perspectives on traditional field communication, IP for field devices, and possible roadmaps to such a future. Their discussions focused primarily on the Advanced Physical Layer, or APL, now being developed in Europe.

The Advanced Physical Layer (APL)

The APL is a technical initiative of several European firms working to develop a technology that would bring Ethernet-like networking to process field devices.  This would include the many devices that require certification because they operate in hazardous environments. NAMUR, the German chemical end user organization, made a series of relevant recommendations for such networks as part of its NE 74 fieldbus document.  These include:

  • Suitability for use in explosive atmospheres
  • Field exchange of devices in hazardous areas, e.g., intrinsically safe
  • Functions of one device do not interfere with the remaining system
  • Robust and easy to use connection technology, if possible by using existing cable infrastructure
  • Flexible network topology, supporting cable lengths up to 1000m and
    spur lengths up to 20m
  • Two-wire cable for both data and device power

process field devicesThe term “physical layer,” or PHY (pronounced “phi”) has a highly specific meaning in network technology. It represents the lowest layer of the OSI network communication model.  A well-specified PHY defines properties, such as physical media type, media properties, geometries, lengths, and materials.  It also defines connector technology, the means of signaling, the carrier frequencies, channel properties, and signal modulation techniques.

No single PHY specification represents “Ethernet.”  Rather, Ethernet networks include many different PHY specifications. In fact, there are over 50 different Ethernet PHYs, almost all of which are specified by the IEEE 802.3 standard.  These Ethernet PHYs employ different media (coax cable, twisted pair cable, single and dual mode optical fiber, plastic optical fiber, etc.) and have different data rates (from 3MB/second to 100 GB/sec) and different segment lengths (from 10 meters to many kilometers).

In the marketplace, the vast majority of Ethernet interfaces shipped today use a PHY known as 100BASE-TX.  This PHY uses a 100MB data rate over twisted pair cable.  All Ethernet PHYs can interoperate because they present a common data link layer.  Thus the question “What is Ethernet?” has a single best answer:  Ethernet is any network that offers an IEEE 802.3 compliant data link layer (see figure above).

IEEE vs. Industry Standardization

process field devicesThe ideal technical solution for field network standardization would be to develop an IEEE 802.3 PHY specification addressing process field devices.  The IEEE 802 organization develops all the related standards, has a very mature standards development process, and has the size to easily manage an additional PHY effort. External organizations are already in place to test and certify 802.3 compliance and interoperability.

While the best technical solution, ARC Advisory Group believes this path would face commercial problems.  Any new PHY requires the backing of several Ethernet silicon providers.  These would have to support the standardization effort, not just by doing the work, but also by process field devicesattending the related quarterly IEEE 802 meetings for two to three years while the work is being done.  By IEEE 802 rules, consistent meeting attendance is a key criterion for determining voting rights on any standard during development.

The out-of-pocket cost for obtaining a single such IEEE vote is roughly $1 million per firm, not to mention the opportunity costs of this effort.  Because of this cost, suppliers are hesitant to back any effort unless they believe there is a large enough market to deliver sufficient return on these investments.  As a result, suppliers use IEEE 802 to standardize features that will apply to very broad markets rather than market niches. 

For example, the Time-Sensitive Networking (TSN) standards are broad enough for potential adoption in automotive, professional audio/video, and industrial automation markets.  The potential market size of an intrinsically safe Ethernet is much smaller; hence less attractive to semiconductorprocess field devices suppliers. These suppliers look to standardize interfaces that will have a market size of millions or tens of millions of units per year.  The complete installed base of process field devices is of the same order of magnitude (roughly 70 million).  Thus in the supplier’s view, the value of standardization is too low to be worth the effort.

The alternative method of standardization is by a dedicated industrial consortium.  This starts with an existing Ethernet PHY, and then involves standardizing an adaptation that meets the requirements of hazardous environments.  This work can be done by an industrial consortium and is the path that has been pursued for the APL. In theory, this work can proceed faster than an international standards development organization (SDO).  However, this path gives up the economies of scale and leverage that come from IT markets and IT/SDO methods. An industrial consortium must define its own rules and practices, and must manage the competitive forces that stand in the way of converging on a single solution.  It also must find its own way to certify compliance and demonstrate interoperability.

Recommendations

  • In the short-to-mid-term, process manufacturers should plan for growing use of IEEE 802.3 standard Ethernet in field devices, but only in applications for non-hazardous environments.  This will restrict such use primarily to pharmaceutical, food, and beverage applications.
  • Device suppliers should look to automotive Ethernet technologies as a starting point for industrial standardization.  These have long lifecycle requirements and demand industrial ruggedness, though they differ from process industry applications in other ways.
  • Both end users and suppliers should weigh the benefits of IEEE 802.3 standardization and look for technologies that could address both automotive and process industry applications.
  • The vision of easily commissioning a field device containing its own DTM or FDI package is compelling.  However, legacy fieldbus networks will persist for years, making this vision unrealizable for now.  Present-day internet technologies will need to be used to identify and obtain the correct device packages for any field device. The solution should be developed and supported by the responsible organizations; FDT and the FieldComm Group.

 

ARC Advisory Group clients can view the complete report at ARC Client Portal   

If you would like to buy this report or obtain information about how to become a client, please Contact Us

Keywords: Ethernet, Fieldbus, FDI, FDT, Field Devices, FOUNDATION Fieldbus, HART, Physical Layer, Profibus PA, ARC Advisory Group.

Engage with ARC Advisory Group

Representative End User Clients
Representative Automation Clients
Representative Software Clients