1 May 2020

The back story to Smart Instrumentation

Industry has always strived to obtain more diagnostics from instrumentation, and this is not a new approach that is associated with the Internet of Things (IoT). 

The evolution of Smart Instrumentation

Before the common 4-20mA signal, basic 0-10Volts and 0-20mA signals were common place, adequately good enough at supplying the DCS or PLC with the process value from the instrument. However, both had one common fault, you couldn’t distinguish between an instrument issue and an open circuit on the cable connection to the device.  So, the 4-20mA signal was born, with defined error currents that are displayed if an instrument failed, and with 0mA being an open circuit. The 4-20mA signal was widely accepted with an increasing amount of installations requiring some form of safety. The new mA “failsafe” value could be interpreted by the PLC or DCS increasing plant / factory safety. To this day, 4-20mA signals are still used on the vast majority of process instrumentation applications.

In the 1980s we started to see the introduction of integrated processors into instrumentation, moving from a complete analogue system to a smarter device. This was the evolution of the smart instrumentation; more intelligence and diagnostic capability, much improved human interfaces and in some cases communication ports to allow connection to configuration software.    

As more instruments became “smart” there was an increasing requirement for different connecting cables and multiple software platforms. However, we still had an issue with communicating to loop powered 4-20mA devices. In 1980, the Bell 202 Modem Standard was adopted as the communication standard for subsea oil and gas. This protocol uses a high frequency half duplex communications to establish the binary values required, with 1200Hz = Logic 0, and 2200 Hz = Logic 1. The Bell 202 signal was imposed on the common 4-20mA signal, and this was called the Highway Addressable Remote Transducer, or HART.   

So, the smart instrument has arrived, with HART being utilised with standard tools to configure devices and analyse faults. However, a connection to the instrumentation wires had to be established, leaving most intelligent process values and diagnostics stranded in the field. This is still the case today with only 3% of the global 80 million HART devices utilising more than the 4-20mA signal.  

An approach had to be developed to establish a remote connection to intelligent instrumentation, and in 1989 Profibus was born deriving from a project that started in Germany in 1986 and for which 21 companies, (including Siemens) and institutes devised a master project plan called “fieldbus”. The goal was to implement and spread the use of a bit-serial field bus based on the basic requirements of the field device interfaces. Profibus Process Automation Bus, (Profibus PA) was established for connectivity to instrumentation using a similar approach to HART, so the instrument was still loop powered.  However, the mA value was fixed and no longer utilised; the process values and all associated diagnostics are transmitted down the same two wires, a true digital signal.  This is more commonly referred to as multi drop, giving birth to a network philosophy.   

Fieldbus Devices and new challenges they present

New technology breeds new challenges and Profibus PA didn’t avoid its fair share of bad installations with a shift away from this standard being seen in instrumentation sales, with only 5% of the two wire devices being sold globally.  So why is this, given this should be the answer to all our digitalisation challenges for instrumentation?  I think that the main reasons for the lack of uptake can be associated with the following: 

  1. Skilled workforce. 4-20mA dates to the 1970’s and has been widely accepted as a connection to instrumentation.  HART communicators have been utilised to configure basic instrument parameters, and then disconnected; hence stranding the digital values.  Now most basic site problems can be solved with a multi meter and a screwdriver. That is until you must replace the device.  More time is now required to fix the problem with local knowledge being key. Ask yourself, could I have established early indication of site problems by monitoring field instrument diagnostics more closely. 
  2. Moving from point to point to network infrastructure. This has undoubtably contributed to the fear and possible lack of uptake for digital instrumentation, such as Profibus PA.  Early implementation saw bad network design that often got amplified over time as more smart instruments where added to the network. While Profibus DP was adequate for most automation tasks, it did present another challenge when implementing Profibus PA, as now there are two networks to design and maintain. The bandwidth problem was also found to be an issue as digital instruments send 5 bytes of data for each process value, with multiple variable devices having up to 15 values to transmit, so 75 bytes from one device. Some Profibus DP to PA converters are limited to 256 bytes so now there is a bottle neck to enable digitalisation. 

Over the next two blogs we investigate how the intelligence from smart instruments can be unlocked for existing installations and new greenfield applications using Profinet connectivity.  When accessing this data, consideration must be made to not increase the complexity of the automation design.  Digitalisation may be in the future for your plant(s) but sacrifices at this stage could be costly further down the line.   

For more information, check out my free online video tutorials here 

Related Tags