Dear All,

There are so many articles on the pros and cons of adopting a field bus communication system. Profibus and Foundation Fieldbus are the most talked about versus the conventional 4-20mA HART powered system.

Especially when it comes to the buses, everybody tries to convince others that his proposed bus is better and brings so many reasons. Each bus has its own fans, which technically are not low in profile as well.

There is a STP to be constructed. The control system has been dictated to be a DCS (no chance for SCADA!), and we are deciding the type of field communication network.

There is a large number of Discrete I/O, say for the limit switches of the valves or Start/Stop commands to the pumps, etc.

There is quiet a large number of Analog I/O as well and the plant is a rather vast, i.e. units are physically distributed.

The question is, shall we go for a Profibus PA network along with a Profibus DP or shall we select a Foundation Fieldbus or no, just have the conventional 4-20 mA system with HART?

Selecting a bus, I think the saving made by fewer cables, JBs, cabinets, I/O cards, etc. will not cover the money involved in more expensive instruments. The point is all the equipment in the plant should be bus compatible, e.g. a normal limit switch.

Should a bus be selected, I am more inclined to choose Profibus as the devices are a little less expensive compared to that of Foundation Fieldbus and also, there are so many Vendors for the STP which are European, so Profibus is more convenient for them as well. To communicate with other vendor packages I guess Modbus would be a good choice and of course there will be an Ethernet for non critical communication issues as well.

For selecting the right bus for communication in the field, I need a clear "detailed FINANCIAL view" of the alternatives in terms of:

- Initial investment for the hardware (All inclusive, i.e. the instruments, cables, JBs, earth work, trays, labor cost, etc.)

- Running operational cost (Operation, troubleshooting, spare parts, technical support cost, etc.)

I need expert comments on this subject with some detailed "indicative financial figures".

Thanks in advance,
Hamid

 

Whew, DCS & a fieldbus for sewage treatment? Not common here in the USA, where SCADA is common.

The field buses were iniated and designed for industries like automotive and refining: intensive capital investment with high dollar (money) value production where a disruption of production loses lots of money. So the value of diagnostics increases dramatically because good diagnostics decreases troubleshooting time and effort. The corollary is that these types of industries hire very well trained specialists who can exploit the diagnostics in field bus instrumentation.

Field buses run on engineered drop-node, trunk & branch type wiring. That's great for batch chemical processors where the batch concentrates dozens of pieces of instrumentation so they can be wired in their particular bus fashion.

But a sewage plant does not process high value material, nor is the process concentrated in one place, and unless your sewage plants hire different personalities than ours, the people on staff are not highly trained instrumentation people ready to scroll through diagnostic codes, they're operators and a supervisor.

I don't begrudge the refineries, batch processors, paper plants, pharmaceuticals or those food processors who use field buses, but a sewage plant?

Stick with 4-20 with HART. It works, and it will serve you and the end-user very well. And shortly you'll be able to use wireless HART with those really remote transmitters (HART wireless standard was announced a couple days ago), so you don't have to do long home run wiring.

Carl

 

FF has some advantages over the other:

. Digital closed loop control (HART does not do that)
. Real-time and non-real-time channels (PA does not separate this)
. Scheduled periodic and synchronized real-time (this gives FF an edge in PID loops)
. Automatic addressing (this makes FF plug and play - no manual addressing. No mistakes)
. Peer-to-peer (this enables control in the field for many loops, this also gives an edge for PID)
. Publisher/subscriber (this enables control in the field for many loops, this also gives an edge for PID)
. Clock synchronization (time-stamping of events at source)
. Alarm & event reporting (tells you when there is a problem without asking)
. Firmware download (unique ability to upgrade devices online, bumplessly, without removing them)
. Robust communication (higher amplitude and longer CRC than HART)
. Powerful and innovative devices (pump health, field display, temperature mux etc. translate into new kinds of savings
. Rigorous interoperability testing (including the EDDL file)
. Status propagation (gives FF an edge in closed loop control)

To learn more about fieldbus and Profibus and stuff you did not know about HART take a look at the yellow book "Fieldbuses for Process Control:
Engineering, Operation, and Maintenance" buy online:
http://www.isa.org/fieldbuses

Cheers,
Jonas

 

DCS is a little over-kill for a water treatment plant. Normally, PLC/RTU + SCADA are used.

In case you would like to go with Foundation Fieldbus, you can look at the RTU from Yokogawa (Stardom), which provides a lot of interesting features additional to the FF interface.

Another options (widely used) is to used 4-20mA+HART. Now that the HART wireless standard has been announced, it is only a matter of time that you can get devices in a cost-effective price range.

Thanks,
Shahid

 

Thanks Carl, Jonas and Shahid,

Actually I am aware of the technical benefits of the BUS system but I need some FINANCIAL FIGURES regarding the saving in the cabling (including the cables, JBs, man hours, etc.) vs. the higher initial investment for the more expensive hardware. Anybody any idea?

Thanks
Hamid

 

I would suggest to go 4-20 mA with HART. And don't forget to take an AMS (Asset Management Server) with your DCS. This will be helpful to do calibration, testing and diagnostic from the control room without going to the field and will be a real cost saver for maintenance. This will be a simple straight forward approach.

Profibus or FF both will do your job. But considering limited instruments in each fieldbus node, special JBs, distance limitations, costly instruments etc. ultimate savings may not be very much.

 

People who make the fieldbus wiring components have those calculations.

Look at
http://www.stonel.com/newpages/protocols/fieldbus/index.htm
for an example.

Carl

 

It looks like you are only looking at CAPEX (capital expenditure = cable and manpower) savings. It depends heavily on the application so one cannot give any general numbers that are meaningful. Major impact on the savings is the distance between process equipment and control room, number of devices, hazardous area requirements, and labor cost at site.

Contact your automation supplier and ask them to make a QBP (Quantified Business Proposal).

Also look at OPEX (operational expenditure = maintenance cost) savings and output (throughput, quality, availability) improvement.

Cheers,
Jonas

 

Intelligent device management software as part of an asset management solution (AMS) is certainly a good idea. Make sure it is build upon the EDDL technology (www.eddl.org). It is the only international (and Euro and US) standard for device integration. EDDL is an integral part of HART, Foundation fieldbus, and PROFIBUS as well as the new WirelessHART somebody mentioned so you can pick either protocol, or you can mix the four protocols if required and still manage all the devices from the same intelligent device management software. For example, you may have radar level transmitters and control valves on FOUNDATION fieldbus, variable speed drives and electrical actuators using PROFIBUS-DP, safety pressure and shutdown valves on HART, and some remote level monitoring on WirelessHART. EDDL works just as well on a Windows workstation in the control room as on a rugged embedded handheld field communicator for field use.

EDDL unleashes the data in simple as well as sophisticated devices: identification, general information, advanced diagnostics, and permits advanced setup, calibration, monitoring, and simulation etc. You get full access and support for device functionality.

Over the past 15 years EDDL (aka DD) has proven very reliable and robust because it is a textual description exactly like HTML (both are descendants of SGML), not executable software. For the same reason it is also secure. Other benefits from being descriptive rather than executable is device version conflicts are avoided (no DLL-hell) and the device files are not affected by operating system upgrades or service packs. Because graphics is generated in much the same way as a web browser generates a web page, the devices will be operated with a very high degree of consistency, regardless of protocol, manufacturer or device type. No licenses need to be purchased or managed.

Cheers,
Jonas

 

The recent web issue of Control Global focuses on fieldbus. There are about a dozen articles on various aspects.

http://www.controlglobal.com/articles/2007/344.html

 

Good day Hamid,

A digital bus would definitely reduces the cost of installation, and commisioning as you are already aware, hence you ask which fieldbus. As we start implementing more digital bus systems we have reduced and improved the on installation methods and improved availability. Fieldbuses go "wrong" in most sites when the installation is done incorrectly, this becomes a costly affair. Training is imperative for whichever fieldbus you use to ensure proper installation and maintenance is performed.

Both FF and PROFIBUS have an excellent set of guidelines that need to be fulfilled to ensure a proper/problem-free installation. This means that for both FF and PA installation costs should be exactly the same.

However, for a more cost effective solution, I have had excellent experiences on PROFIBUS over the last 6 years, and I believe it is definitely a better option and a more cost effective solution. If you do not need "control in the field" then PA is definitely the way to go.

As an example, the PROFIBUS Middle East organisation has just helped commission and test a site installation where 21 PA devices were connected to a DP/PA link (in total 7 PA segments with just over 80 devices in total, and just under 5 days to test and communicate to them all). The cycle time was 353ms per cycle on the largest network (21 devices).

What has been impressive is the amount of devices on the network and the no. of various vendors on the same network (7 different vendors) that was installed without any difficulties in their interoperability. We also tested the interchangeability of some of the devices and it works excellent (please use PA Profile 3 (GSD) to ensure openness).

With using a complete PROFIBUS system your DISCRETE and PROCESS world is one common platform, one engineering std, and training, support, fault finding and maintenance becomes easier to manage. So be it a VSD, an Overload, a power meter, a softstarter, an ACB, weighing, digital I/O stations, all instrumentation from any vendor of your choice... you would be able to configure your network easily without any difficulty.

Using multiple fieldbus technologies ends up with interfaces and different ways of handling devices/engineering (cumbersome and most times expensive).

HART devices can also be integrated into the PROFIBUS world, but I have to warn you, that this may work out cost effective but the delay in the unscheduled or acyclic information i.e. configuration and diagnostic info. is a bit on the slow side. But again it is still quicker than going to the device in the field.

In short, more devices on a segment means more cost savings, and for a STP cycle time of 350ms is more than sufficient. A SCADA or DCS can be used on your site, with a SCADA most probably proving perhaps a bit more cost effective.

From an installation cost point of view, both FF and PA should be almost exactly the same, with price deviations on the actual instruments. Depending on how you maximise the no. devices you add per H1 or PA link, PA links would be fewer.

I invite you to visit one of our multivendor installation sites in the Middle East region and check with the end users on site what they have to say about the cost of engineering, implementation and maintenance.

Regards,
Pregs Naidoo
[email protected]

 

In my personal opinion PROFIBUS-DP is excellent for drives and motor starters. When it comes to discrete signals such as from switches and to solenoids these do not connect directly to PROFIBUS-DP. You could connect them to remote-I/O using PROFIBUS-DP. That is, PROFIBUS-DP is more of an intermediate level remote-I/O bus than a low lever discrete bus. To connect discrete switches and solenoids to a bus you would need to use DeviceNet or ASI.

Without getting too technical (I will try) I just like to take this opportunity to mention that the PROFIBUS bus cycle is very different from the FF macrocycle. The PROFIBUS bus cycle is only about acyclic communication. The time does not include measurement, control, and actuation. The actual loop time will be longer than the bus cycle. The FF macrocycle includes cyclic and acyclic communication as well as measurement, control, and actuation - the whole loop time. It therefore may appear longer. The FF macrocycle has the additional advantage of being precisely periodic which is important for PID loops and process control functions in general. However, it does not matter for motor control or discrete applications where PROFIBUS works just as well.

So FF H1 for PID loops, PROFIBUS-DP for motor control, and ASI/DeviceNet for on/off devices complement each other very well. EDDL (www.eddl.org) makes it easy to manage FF and PROFIBUS from a common tool. I personally wish you could use one single bus technology to do it all, it would indeed be easier, but you can't. I have been an advocate of that for a long time and I still am. There are no variable speed drives using FF. There are no proximity switches using PROFIBUS. Integrating the data coming over different protocols using EDDL is currently the best available option. EDDL also integrates any wired HART and WirelessHART into the same software.

Cheers,

Jonas