1 July 2005

Think Digital

Integration of fieldbus without a DCS.

By Frank Jacobs and David W. Appleby Jr.

It was early in 2004, and Matthews, N.C.-based engineering firm Perigon needed to commission a major bus-based system for an international polymer resins and urethanes coatings manufacturer.

The products produced in this system would see use in hardwood floors, automotive supplies, and as ingredients for Sherwin Williams' products.

The direct customer was a Germany-based company looking to establish U.S.-based manufacturing facilities.

Perigon served as the architect and turnkey system supplier for the entire process automation system (PAS) portion of the overall project. Therefore, the company was responsible for the system vendor selection. Because the project would involve smart field devices, Perigon also was responsible for fieldbus selection and implementation.

The end user originally specified a German automation controls company for their system because their existing European facilities incorporated their products. However, the client's need for local support within the U.S., coupled with a desire to leverage newer bus-based technology, primarily for its cost-savings and simplified startup advantages, ultimately influenced their final decision.

Due to economic conditions, however, the project evaluation process spread out over a three-year period. This provided an unexpected benefit in that Perigon was able to review new and emerging technologies in addition to considering advances taking place in non-DCS supervisory controllers.

Jump on the bus

From the beginning, everyone decided on using open digital bus technology for the system's device level. The requirement for a bus-based batch/process control system was due to the owner's desire to install a system that also would meet the needs of future applications. The selected all-digital network also would need to provide value throughout the system's life cycle, including:

Ease of installation
Reduced engineering, installation, and commissioning costs
Improved system operation capabilities through better resolution and accuracy (This would come through the use of digital technology versus analog.)
Improved device awareness and maintenance through improved remote diagnostics (signal and calibration) and remote-configuration capability

Bus or digital technology was the best way to achieve this because of its ability to provide more robust instrumentation and control devices. Digital devices would provide better process information in the form of higher resolution, in addition to better accuracy through the elimination of analog signals. Digital buses also would bring the ability to remotely configure and diagnose the instruments, leading to faster commissioning and better ongoing system maintenance.

The original concept was for a single network to support analog and discrete field mounted measurement and control devices. But, as system field instrumentation selection progressed, the team noticed there were considerable differences between the operational requirements at the device level. These differences come with the application requirements typically identified with industrial control (highly discrete) and process control (continuous), such that the different types of networks address a specific set of control and operational requirements.

In the end, the companies selected two device networks, since using one type would not provide the full value desired across the automation spectrum.

Once it was determined the system would require both types of networks, the hunt was on for a process automation system that would support multiple control disciplines as well as both network types.

PAS requirements

The distributed design of the manufacturing environment, including multiple floors and independent unit operations, established base system requirements, such as:

Capability to provide integrated process, batch, and discrete (industrial) control.
Ability to manage multiple batches while coordinating multiple trains.
Single operator interface that you could use across the facility capable of enabling the operator to see and interact with all controller or device data.
Integrated support for multiple device networks with a high degree of scalability.
Reducing on-site maintenance in an effort to minimize overhead costs.
Strong local control system and network equipment support (for both installation and ongoing facility maintenance).

After establishing the requirements, they had to perform an in-depth evaluation of device networks and control systems in order to identify those that could support the automation requirements.

Considered were standard device bus technologies such as Profi-bus, DeviceNet, Asi, and Foundation Fieldbus, as well as systems from three global companies.

As originally defined, the application required a DCS-based solution, primarily due to its process content. It was felt that Foundation Fieldbus required a DCS. However, as the evaluation process continued, due to a decline in economic conditions, the process design changed, and the size of the initial application shrank in scope. This change added a new dimension to technology, vendor evaluation, and system scalability, in addition to a reevaluation of the control system requirements.

Scalability became even more critical in an effort to meet the customer's requirements to implement several batch trains, in addition to being able to easily duplicate these trains at a later time.

In the end, the companies decided utilizing a DCS was too costly for the initial applications; however, the system still required full batch/process and industrial-control capabilities as originally defined, including the FF interface, while remaining within the limited budget.

In the final analysis, the solution would utilize Foundation Fieldbus for the process instrumentation and DeviceNet for the industrial control. Ethernet HSE/IP and ControlNet would also be the control and information backbones. Independent linking devices that could go through the facility could also bridge the H1 directly to the Ethernet backbone to connect the FF devices to the process automation controller. The solution required having the controller support all of the multi-disciplined control languages, in addition to integrating with the device networks. This allowed Perigon to provide a solution with a simplified system connection architecture.

Foundation Fieldbus enabled distribution of the process devices as well as control throughout the system and into field devices (instrumentation) if required. (While they would not use this capability in the initial installation, it was a definite plus and could be used in the event that backup and local control would be added later.) DeviceNet then provided the support for the Industrial control devices such as discrete valves, drives, and tank scales. Both of these networks supported critical control devices and would therefore require a tight coupling with the supervisory controller.

System under control

The resulting process control system solution comprises only bus-type devices that control 12 storage vessels, a reactor, two mix tanks, a cooling tank, a holding tank, and two other mix tanks. A total of 58 process instruments from multiple vendors provide 88 Foundation Fieldbus analog channels.

DeviceNet based motor control centers with motor starters and drives complete the system.

The system incorporates supervisory batch software, PC-based operator consoles, a multi-disciplined controller, DeviceNet bridges, and FF linking devices. This combination of system equipment provides flexibility and scalability that is not normally available in a DCS, yet still provides for full integration of the bus networks.

This approach provided some unexpected capabilities, especially on the FF side. The control equipment vendor offered a standalone, multi-segment FF linking device with the ability to distribute the H1 interface throughout the facility, while connecting them to the supervisory controller through Ethernet (HSE the Foundation Fieldbus High Speed Ethernet and ENet/IP protocol). Together, with the process automation controller, the linking devices provided additional flexibility and costs savings. This also left the door open to easily add devices and, if required, control loops, to these field devices at no extra cost.

As a practical approach to the network design and implementation, management of the documentation of the field device networks was separate from the control P&IDs. By separating management, it greatly simplified design reviews since control system evaluation became a review of the logical view of the system, rather than having to simultaneously take a look at the control system's physical implementation.

While you could represent both models (logical control and physical control) on the same drawings, there was enough process piping and enough control without having to deal with the potential confusion from other layers of detail communicated in the same set of drawings.

An unexpected benefit occurred when the engineering team realized when they had to reallocate field devices to different field networks, there was no real impact on the control drawings. This helped mitigate excessive costs and time delays that would otherwise require having to review more progressively complicated drawings.

Perigon also addressed the protection from high and medium voltage systems by installing the Foundation Fieldbus cabling in conduit between the control panel and P&F field barrier devices. This occurred in classified areas and through conduit out to the Interlink BT Fieldbus distribution bricks. This is the primary area where the power and communication cabling route in close proximity to each other.

Learning experience

Although Perigon had prior experience with DeviceNet, this was the company's first application utilizing FF devices.

There are similarities between DeviceNet and Foundation Fieldbus, at least as far as the network. The biggest difference was in the interoperation of Fieldbus devices and the potential to configure control strategies in the devices. This is a feature not used in the initial application but is something to consider for the future.

The following are lessons learned:

Instrument selection becomes more critical to the operation of the system. When the only signal available was an analog 4-20mA, the system engineer and instrumentation engineer did not need to coordinate, as the only common element was the tag number. Coordination is critical with digital instruments in order to ensure they get the proper instrument for the process measurement, as well as for knowing the device will support the desired control strategy/system operation. This requires an understanding of the process measurement aspects and the Fieldbus device side. The added capabilities to devices can (and do) change between device vendors.
Some devices support a full set of function blocks, which is useful if control in the device is a possibility. Others only provide for a few. This is important based on the systems control strategy.
If costs are important, some device vendors provide FF, but only at an additional cost.
The ability to remotely locate the H1 interfaces (Linking Device) throughout the facility reduces the required engineering.
Selection of a PAS supplier who will provide connectivity to the FF devices and qualification support for the FF devices, which are proposed and supplied on the project.

Bus benefits

Foundation Fieldbus is an emerging technology. As such, it is critical to recognize that, as with any new technology, there is a learning curve. You should build in that learning curve into any first-time project. However, once completed and the technology is understood, including capabilities and limitations, there are some very clear benefits. These include:

Communication between the multi-disciplined process automation controller and the field devices is a producer-consumer interaction rather than a master-slave interaction.

PAS provided flexibility in the connection to the FF devices. This allowed expansion in the number of FF devices during the course of the project.

Ability to view networks as a virtual network is available from both network types, which enables the "drilling down" from an upper level workstation (local or remote to the facility), thereby allowing the opportunity to monitor any device on either bus for diagnostic information.

Ease of installation occurs through the use of prefabricated cables. These cables are more expensive but reduce the time for installation and commission such that they actually provide a savings.

It's important to note all medium and high voltage sources come encased in conduit using explosion-proof methodology through to their destination. Shielded cable also provides all fieldbus segment wiring with proper grounding techniques. The preferred approach is to always try and route power and instrumentation/communication wiring with as much separation as possible in order to avoid interference problems. This step is relatively easy to do when performing a new installation.

Diagnostic capabilities are available built-in and remote to the devices and are completely accessible via the vertical network. You can monitor device operation from a dedicated workstation, a local operator panel or workstation—external to control environment—or remotely from the site. This was particularly helpful during field implementation and startup.

Ability to embed control at the device. Although this was not used in the initial application, having this capability allows for future application extensions.

Ability to have strong data and application constructs at the field device, including analog, discrete, loop and multi-state devices.

Dramatic savings in cost of implementation include material costs for conduit and labor related to the distributed instrumentation and control systems was reduced while maintaining a coherent, integrated environment for the PAS.

Reduced cost of change management realized savings when they redistributed field devices during the design process; they also saved with installation configuration, and post installation field device management.

Once you understand how FF works and work with a few devices of different varieties, any new devices are very easy to understand and work with. In the end, you don't have to own a DCS to take advantage of FF. It is very practical and equally beneficial to implement FF within what would traditionally be a "hybrid control system" environment.

The critical issue is to understand and accept you are dealing with a digital network. If you approach the application as if it is a 4-20mA signal, you will have trouble. A bus-based application requires it must be treated and approached as such. In addition, you should select a PAS supplier that has the ability to support not only his control system products, but who can also provide initial support for the bus-based devices. This was a big help in resolving any interoperability issues.

Behind the byline

Frank Jacobs was the lead technologist and lead project engineer for the design, vendor selection, system implementation, commissioning, and on-going system management for Perigon. David W. Appleby Jr. is in product marketing for Process Control Platforms at Rockwell Automation.