The intriguing case of fieldbus fragmentation

Chances are that you are using either an IOS based mobile phone or and Android based phone.

If you are using a IOS phone  you are most likely using it running the latest OS version (9.x.x....) but if you are using an Android phone you may be running anything from 4.2 to 6.0 OS version.

This issue is usually called Android fragmentation, and is currently becoming a developers nightmare. 

The availability of Android OS upgrades is limited due to the fact that it depends of Google developers, manufacturers willingness to make the updates available to their devices and in many cases the carrier's willingness to make those updates available (this last factor less relevant lately due to the growing tendency to use free unlocked devices).

Of course this is also related to the wide variety of hardware combinations that Android phones use.

 

Fragmentation in Fieldbus technology

IEC 61158-2 technology seems to have gone in a similar path. Initially Foundation Fieldbus and Profibus PA were supposed to be highly flexible technologies with multi vendor support and a great level of inter-operability.

At the field device level this was basically accomplished, but at the infrastructure level the fragmentation issue has become present in a way that has affected the acceptance of this technology in the last years.

 

What is Fieldbus infrastructure?

The fieldbus infrastructure components are basically the following:

1. Fieldbus power supplies (a.k.a. power conditioners): these devices provide energy to the segment while allowing MBP coded communication through a 2 wire twisted cable. They usually feature galvanic isolation.

2. Device couplers (a.k.a. distribution blocks, segment protectors, megablocks, etc): these devices isolate a short circuit located in a spur from the segment, thus avoiding segment communication loss.

3. Surge protectors: they protect the infrastructure components from damage due to surge events.

4. Field Junction Boxes: employed for the montage of device couplers and surge protectors in the field.

5. Fieldbus cables: 2 wire twisted pair cable, with specific electric characteristics.

6. And for Profibus PA, segment couplers (a.k.a. links, linking devices, etc.)

 

The evolution of fieldbus topologies

You can access an interactive summary of the topologies mentioned in this article here.

 

Fieldbus in safe areas

For a generic application in a non Hazardous area the solution is straightforward: Use a standard fieldbus power supply, fieldbus type A cable, and the device couplers of your preference. Add some surge protectors if lightnings, transient or surges are an issue  in your case, then you are basically done.

 

Fieldbus in hazardous areas

But the main argument for the use of fieldbus technology is the simplified wiring for hazardous area applications. In fact I've found that many end users have a tendency of considering both FF and PB PA an Ex-Zone specific solution.

Here is where things become funny:

You can install fieldbus using the Ex-d protection method, which allows you to reach Zone 0 with your field devices and lets you mount your device couplers up to Zone 1. If you live in the NEC side of the planet, you'll be installing your field devices and device couplers in Class I Div 1 areas. Of course the device couplers shall be equipped with the corresponding Ex-d housings.

But you′ll loose quite a few benefits of fieldbus following this path, like live replacement of the field devices or the need of a hot work permit for any maintenance requirement, to name a few.

Fieldbus was designed to be compatible with the intrinsic safety protection method. As you surely know, explosion protection by intrinsic safety is based on the concept of employing energy limited circuits in the field. This concept requires the use of intrinsically safe field devices and energy limiting interfaces (zener barriers and more frequently in later years intrinsically safe galvanic isolators).

 

Intrinsically safe Fieldbus

In a IS fieldbus, the power supplies are the IS barriers.

The entity concept: The first approach was to employ the Entity concept: each component of the IS circuit is considered as a different entity in the calculations required for the IS proof. If you had the bad luck of being an early adopter of IS fieldbus, you would have had to deal with Entity power supplies, which limited voltage to 10,5 Volts DC and limited current to 70 mA.

An Entity fieldbus segment would let you use 3 to 4 field devices if you were lucky.

The FISCO concept: FISCO was an empiric approach to solve Entity limitations developed by the German PTB (Physikalisch-Technische Bundesanstalt), the mother of all industrial testing institutes of the world.

They found that if you employed an specific kind of cable (fieldbus Type A), limited the cable length to 1000 m, limited the spurs to 60 m and employed FISCO certified field devices, you could increment voltage to 12,5 volts and current to 110 mA in IIC applications (280 mA in IIB, but no German engineer would design for anything else than IIC).

This allowed you to basically double the number of field devices per segment, making IS fieldbus economically viable and competitive. But you had to use FISCO certified power supplies, which were bigger, hotter and more expensive than standard ones. It also prevented the use of redundant power supplies.

 

The High Power Trunk

A few months later, an alternative installation concept was presented: the High Power Trunk concept or HPT.

This method showed how you can use the standards restrictions for your benefit:  ATEX standards allow the use of non arcing or non incendive cable installation techniques in Zone 1, so the HPT concept moved the energy limiting circuitry from the fieldbus power supply to the device coupler (now named isolated device coupler or fieldbarrier) and using IS just at the spur level.

The isolated device coupler could be mounted in a Ex-e housing in Zone 1 (or in Class 1 Div 2 for NEC fans) and you could use standard fieldbus power supplies, even redundant ones. Now the number of field devices went up to 16 (real life numbers, the required electronics distorted the MBP square wave too much after the fourth fieldbarrier to ensure safe communication).

You just had to install the trunk cable in the appropriate way: basically protecting it mechanically from damages and with the adequate cover for the environment.

And either FISCO or Entity field devices could be connected at the spurs.

The FISCO and HPT concepts lived together competing in the big FF projects of the 2000's, mainly in Asia and the Middle East. And we are talking about BIG projects, featuring thousands of segments. Oil prices were rising until they got well above 100 USD and the ROI of a greenfield O&G related plant in the middle east was something like 8 to 12 months. I guess this was the golden age of fieldbus.

 

Saving money in Zone 2 

But in a typical O&G plant, about 60 to 80 % of the field devices are not located in Zone 1 or 0, but in Zone 2. FISCO and HPT were too expensive for those cases. There was room for innovation.

The FNICO approach: One supplier developed the FNICO concept (FISCO for Zone 2), FNICO power supplies allowed 12,8 V @ 280 mA in Zone 2 (for IIC applications) and 13,1 @ 340 mA (for IIB applications, never mind the German engineers). So you got IS in Zone 2 for trunk and spurs.

HPT for Zone 2: Other supplier developed the HPT for Zone 2 concept: Use a voltage limited power supply, an Ex-e or Ex-na type of cable installation method and a current limiting device coupler. Zone 2 IS was achieved at the spurs.

 

This was the moment when fragmentation started 

In the Profibus PA world people went for FISCO all the way, but in the FF a lot of devices were available only in Entity versions. FNICO never reached the standard level.

 

The Ex-ic standard

The issue was settled by the development of the Ex-ic standard. The tricky part was that if you employed FISCO devices in your Ex-ic segment, you had to use a power supply limited to 17,5 V.

And if you had Entity field devices you had to use a power supply limited to 24 V. There were Entity to FISCO adapters, but they were really expensive. There were also Entity to FISCO connectors. But this approach doesn't look smooth. Some field device suppliers chose to design devices that were certified both as FISCO and Entity.

 

DART comes to town

The creators of the HPT concept developed a technology that allowed increased power in the trunk (up to 24 V dc and 360 mA in Zone 1) by the use of electronics that detected the characteristic current variation over time in a short circuit and then shut down the segment power until the fault was gone. In fact for the duration of the fault the system became strictly IS, afterwards it was reset to the nominal design levels. A very clever approach, but you needed an special power supply and special device couplers. Also this technology was not compatible with FISCO devices, although it allowed redundant power supplies.

 

"Redundant" FISCO

The people that developed FNICO came with a redundant FISCO solution, which was in fact a hot stand by kind of redundancy. A monitoring  module checked continuously if the primary power supply was working OK, and switched to the secondary in the event of a fault. Not as smooth as power balancing redundancy, but certainly a viable solution.

 

Redundant HPT isolated device couplers

The first generation of isolated device couplers were somewhat sensitive to grounding issues. Some end users considered them to be not as bulletproof as promoted. Second generation devices were reliable, but in order to increase reliability and also to offer a alternative in a market where every supplier had its own isolated device coupler version, the redundant FISCO developers designed the redundant isolated device coupler. It was a hot stand by fieldbarrier, with a clever modular design and plenty of configuration options, for a price of course.

 

Solving the single trunk issue

If the choices were not enough, another supplier developed the active or automatic termination technology. The idea was that since the fieldbus technology did not feature the option of a redundant trunk, the trunk was a single point of failure in the event of a wire breakage. In most applications the trunk is carefully mounted with mechanical protection, even in safe area applications, so the chances of wire breakage are really low. But for the very cautious end users, a ring topology was like an extra safety layer.

 

Fieldbus ring topology

Fieldbus ring topology required the use of device couplers with automatic active termination, which are really expensive. It also required hot stand by redundant fieldbus power supplies, also expensive. And limited the cable length to a half in real life, because you had to return to the power supply with the cable in order to close the ring. Also nor the device couplers or the power supplies were compatible with non ring capable versions.

 

DCS suppliers introduce the SMART I/O concept

I guess that at this point DCS suppliers started to get fed up with the task of selecting and specify the correct fieldbus infrastructure components for the application. And they really didn't like to depend on third party suppliers for the field device interfaces. Competition was really hard for the ongoing BIG fieldbus projects, fieldbus infrastructure suppliers went bone deep with discount in order to win bids.

And then one DCS supplier came up with the Smart I/O concept: since DCS suppliers worked very frequently with thousands of I/O points, and quite a few of them required IS protection , they developed a flexible I/O subsystem that allowed the customization (the preferred term was characterization) of each I/O to the corresponding field device.

The I/O system was based on a backplane that was connected to the controller via redundant Ethernet connections and featured individual slots for each I/O point. In those slots you plugged a module that could turn the I/O point into a DI, DO, AI, AO or TI card, included HART support if necessary and could include IS circuitry where it was needed. 

Did you want the advanced diagnostics of Fieldbus? use HART field devices and the appropriate Asset Management tools. HART support was on board, so no need of expensive multiplexers and serial to Ethernet converters. Of course these device diagnostics were not as thorough as the ones available in Fieldbus, but it was more a matter of what was enough.

 

Foundation Fieldbus killer features

Foundation Fieldbus featured a couple of killer features like automatic device addressing, automatic live firmware and software updates and the coolest one: control in the field or CIF.

You could run control loops in the field devices themselves, no host required. You could use CIF in order to lower the controllers processing load or as a backup feature that ensured the control in the event of  a host failure. Unluckily this feature suffered from an urban myth: that it was not commonly used except by the true hard core fieldbus integrators.

 

The truth behind CIF

I've worked with a few EPC companies and DCS suppliers and CIF is commonly used in real life applications.

The coolest feature of this already cool characteristic was that CIF could be applied using field devices in different segments, but only if those segments were connected via HSE or High Speed Ethernet: the Ethernet implementation of Foundation Fieldbus.

Only a couple DCS suppliers employed HSE, and I guess that only the really hard core fieldbus specialists of those companies attempted to use CIF between segments.

 

Why HSE was not adopted by DCS suppliers?

My guess is that DCS suppliers run a services based business, but in their heart they still consider themselves as hardware suppliers, so HSE meant less controllers sold and the possibility of easy controllers inter operability between brands. But this is just a guess. Additionally, most DCS suppliers preferred to keep using their own Ethernet protocols instead of switching to HSE. 

To make things worse for Fieldbus, oil prices went from well above 100 USD to the low 40's USD in the last couple of years. Fewer O&G projects meant rougher competition between DCS suppliers, who then began to see Fieldbus as a expensive extravaganza.

After all, Foundation Fieldbus segments used for control were limited to very few devices. And most of the DCS systems already had available their own version of the Smart I/O approach, now featuring the so called "universal" multi channel I/O cards, each channel configurable by software with integrated IS barriers and even optional on board signal conditioning.

Finally Wireless HART became an international standard. This meant that, at least for monitoring applications, no wiring was needed, not even the two twisted wire, MBP based fieldbus cable. ISA 100 looms around.

This is where  we are just now. I really like fieldbus technology, but for some end users the ROI time is too long. Their I&C guys are not too willing to get deep into CIF, Fieldbus topologies and grounding requirements.

I think this is a pity, because the Fieldbus promise was really awesome. But Fieldbus in its more pure incarnation (H1 segments interconnected via HSE linking devices using CIF all across the plant) only became true in very few cases.

The current trending topic in Automation is the IioT or Industry 4.0 concept, which features pervasive sensing, smart devices automatically addressed and access to huge amounts of diagnostic and process data in the industrial applications. Fieldbus could be the way to deploy IioT in the field, the way for process automation to become fully digital. To leave analog technology in the past. I wonder if this is still possible.

 

The "profiles" approach

The current approach for Fieldbus segment topology design and selection is based on the profiles approach. There is a so called Open Integration initiative led by some of the top industry suppliers. The objective is to make easier the process of topology selection according to the application requirements, the control system used and the availability needs. You design your segments according to the worst condition and by doing this you are sure your design will be able to cope with less demanding requirements.

Segment design software has been available by every major infrastructure supplier, although each app usually is focused on the suppliers own hardware offerings.

 

A new physical layer?

Lately a new approach has appeared: its called Advanced Physical Layer for Process Automation. Its an implementation of Ethernet that also provides power to the field. You go from the controller to the I/O cabinet via the system's Ethernet version, connect a kind of Ethernet switch or coupler that converts standard Ethernet to the Advanced Physical Layer Ethernet or APL (communication and power trough a 2 wire cable, which can be also intrinsically safe) ant then go into the field and reach a compatible device coupler.

This last device coupler has a neat trick in its firmware: you can connect FF devices, Profibus PA devices or HART devices to its spurs, and apparently it will automatically configure itself to convert either one to the APL.

This is really cool, but it also means that if this approach wins, Fieldbus will simply become a way to connect a device to a junction box. And that is really not the original purpose of Fieldbus.

Fieldbus history is a fascinating tale of technology advances and politically induced contradictions. As the first industrial network standard developed by the users for the users, it has come a long way, but it seems to have always been just short of becoming the definitive solution.

Last year the Hart Communication Foundation, the Fieldbus Foundation and the FDI consortium became the Fieldcomm Group organization. It will be interesting to see if this new approach manages to fulfill the original Fieldbus promise.

 

Mirko Torrez Contreras is a freelance Process Automation consultant with a longstanding faith in digital technology. He keeps thinking that fieldbus technology contains great features, but these features have not been adequately promoted.