Is HART 7 set to bring us ANOTHER KILLER APP?
11 March 2009
The stage is set for a major change in the way instruments communicate their values with control and asset management systems. With the debut of HART 7, the possibility is now open to collect signals through both wires and wireless systems, making data collection much simpler to organise and maintain. This may turn out to be the ‘killer application’ of the decade.
E-Senza Technologies' WirelessHART gateway
Over a period of nearly 20 years, the HART protocol has moved into and found its special place deep inside the process automation’s instrument systems.
A hybrid digital-analogue technology born in the early days of the digital revolution, HART’s acceptance was relatively easy, because it amounted to nothing more than a modest telephone signal imposed on top of the already ubiquitous 4-20 mA direct current conductor for transmitting the process variable. According to the HART scheme, the instrument could continue sending its process variable as before, while on top of it, when requested, it could also transmit some basic identification, configuration, and diagnostic information in a new digital language that HART communicators could readily understand.
The new addition didn’t interfere with anything that was already there, didn’t require any additional wiring, and provided one immediate benefit that almost everybody liked at once: a simple way to configure all new flow, pressure, and temperature instruments, regardless of which vendor they came from.
This isn’t what the HART visionaries had in mind when they first created the protocol, but it was such a success they could hardly deny it. Soon virtually all industrial instruments were manufactured as HART devices, because the additional electronics, in high volume, became a trivial fraction of the cost. The benefit for the instrument technician was that he could unpack any device from any vendor, set it up on his work bench, attach a HART communicator, and configure it for operation. Thus millions of devices went through this HART initiation ritual, but once retired to the field they never fulfilled a HART request again—unless the technician brought his configurator out to the field to make an adjustment.
The overwhelming success at the benchtop didn’t alter the HART fieldbus vision many of the creative engineers had. They kept their eyes fixed on the benefits full HART communications would bring, not just to an instrument configurator, but to the entire plant control and management system.
Time and again, over the years, they issued updates to the protocol to enhance its communications performance, but outside the laboratory benchtop, there wasn’t much communication going on. A few manufacturers built multiplexers to receive HART signals from the instruments and put them directly into their control systems, but these didn’t take the market by storm.
Meanwhile, on the legal side, international standardisation was pursued and won. At the same time, the HART Communications Foundation reached out to the sometimes-rivals, sometimes-friends Profibus and Fieldbus Foundation to secure technology and software co-operation that allow the systems to work together.
QUO VADIS HART 7?
A ‘Killer Application’ is something that propels a new technology to widespread adoption. For example, spreadsheet and word processing software made early PCs readily accessible to the public.
For HART, the killer application in the 1990s was benchtop instrument configuration. If that seems to be an exaggeration, remember industrial controls is painfully slow in adopting new ideas. Terry Molloy is quoted as saying, ‘We use state-of-the-art technology with fifteen to twenty years of proven reliability.’
We may now be setting the stage for a new HART killer app. This one clearly will be brought about by wireless, and it will be largely for asset management systems.
But wireless control is not completely out of the question. Control loops, if they come under HART’s spell at all, are likely to be used in a supervisory or non-critical applications. They will also very likely be relatively ‘slow’ loops. End users have told ABB’s wireless product manager, Gareth Johnston, that initially, it would take between five to ten years before they contemplate closed loop control over wireless. ‘If you consider how many control loops there are using Foundation Fieldbus—not very many—this comes as no surprise,’ he says.
When HART 7 was first released in September, 2007, it was the wireless feature that received the most attention. But version 7 contains significant improvements over HART 5 and 6 for wired instruments, including time stamped data, increased diagnostic capabilities, and the ability to report by exception (see table). In keeping with the long-standing HART philosophy, all of the latest revision is compatible with previous versions. Not only that, but by design wired and wireless HART work together seamlessly as a digital field sensor and control network. The operate in the same way, program the same way, and are completely interchangeable.
Clearly thought, there’s no hiding it: the current rage is for WirelessHART. Nearly 20 companies are busy with soon-to-be-released products, and all of them will interoperate, not only with each other, but with all HART devices ever made—HART’s certification programme will see to that! Hoffer Flow Controls new HRT 1 rate indicator / flow totaliser has the honour of being the first instrument officially certified for HART 7.
As with previous versions of HART, there is a basic command set to which all devices must conform. There is first of all the Universal Command set to perform instrument verification and read variables and diagnostics in the device. There is also a set of Common Practice commands, which manage the device for ranging, loop current set up, and so on. Universal diagnostics is something included in all HART devices, so that the device’s performance is continuously monitored for malfunctions, configuration changes, and variables out of limits.
Smart Reporting (see end of article for explanation)
For wireless operation there is a new set of commands for setting up, operating, diagnosing, and monitoring the performance of the network. These wireless commands define all requirements for joining, security, message routing, diagnostic reports, bandwidth allocation, and other necessary capabilities to ensure interoperability, and proper performance. One of the key values the wireless network continuously monitors is the remaining battery life of each of the sensors.
WHAT WIRELESS WILL BRING
The cost of installation of traditional instruments has been a barrier to many plants’ adoption of more sensors in the field, which would obviously benefit their intelligent manufacturing efforts. It has been estimated that the planning and running of signal cable can run as high as $10,000 per point, when all the factors, such as running conduit and cable trays, drilling through bulkheads are taken into account. Tapping into the pipe of a product stream may in itself be a major operation. Small wonder there is such an increased demand for non-invasive sensors and multivariable transmitters! But wireless installation will have fewer, or ideally, no wires to run.
One immediate benefit is short term or ad-hoc measurements, which now become a more likely scenario due to the intrinsic low cost installation a wireless solution provide. Even when local power is available, there is often not enough signal cable back to the control room. For example, a troublesome heat exchanger may have a blockage at an unknown interior location. Inserting thermocouples into spare thermowells, or temperature sensors mounted temporarily on the surface, might assist in locating the blockage. Once found the sensors are quickly removed.
Then there are the difficult to reach positions in the plant, that are hard to climb up to, and even harder to install permanent wires. High up on a distillation column, a wireless sensor could replace a local gauge and eliminate the difficult cable run.
In the larger picture, the low cost nature of wireless instruments will allow new points to be measured, which will solve some process problems and give plant operators greater visibility of their plant equipment. Much data collection today is still done manually; wireless collection may possibly automate much of this activity.
Even process data that are transmitted over the traditional 4-20 mA analogue lines suffer from the fact that only one variable is permitted at a time. Digital WirelessHART makes multivariable transmission easy.
And then there is the issue of the ‘stranded data’—valuable diagnostic and performance data locked away in wired instruments who stay where they are because they have no place to go.
‘What value is there connecting a wireless network to an existing HART 5 instrument?’ asks Mr. Johnston, rhetorically. ‘A positioner or instrument with moving components or consumable components is an obvious target. By adding a wireless adapter to a positioner, it could still use the analogue signal for control, but with wireless HART the valve position could be read and compared to the setpoint. This along with other diagnostic data heretofore untouched by the control system could lead to a new understand of how the valve works and when it needs maintenance.
THE TECHNOLOGY PROVIDERS
Most industrial instrument makers rely on networking product supplier Dust Networks, who pioneered the TSMP (Time Synchronised Mesh Protocol) concept for wireless mesh networking, and donated the technology to the HART Communications Foundation. Dust is now shipping its SmartMesh IA-510 system, which it calls the industry’s ‘first interoperable embedded WirelessHART solution.’
It combines the Intelligent Networking Platform with low-power radio technology to deliver a complete HART 7.1 embedded solution, and this allows industrial automation vendors to integrate wireless networking into their existing sensors and solution architecture or to update existing devices with wireless adapters.
‘Interoperability has been a missing link in the wireless sensor networking industry and one of the key reasons end-users were reluctant to invest significantly in earlier technology,’ said Dust Networks’ president and CEO Joy Weiss. ‘Major suppliers are now offering interoperable wireless sensors. This is an industry game-changer that, combined with the economic benefits wireless solutions provide, will lead to ubiquitous use by the industrial market,’ she added.
Wireless adapters may in themselves form a sub-industry for WirelessHART, given that there are already 26 million wired HART devices in the worldwide installed base. Adapters are devices which plug into an existing HART instrument to pass the data through a WirelessHART network to the host. The adapter could be located anywhere along the instrument 4-20mA cable (a significant advantage, given that some locations may be better for radio transmission than others), it could be battery powered or it could obtain its power from the 4-20 mA cable. Some adapters will be battery powered and use the same battery to power the instrument as well—in this case there will be no need for an analogue signal to the host; all process data will be sent via WirelessHART.
WirelessHART System Architecture (see end of article for explanation)
Most would agree that adding a wireless adapter is a lower risk option than rewiring a marshalling cabinet to add a HART multiplexer. It also allows end users to add the wireless adapter point by point, where and when it is required.
But, first things first. Before we talk about major retrofitting, industry needs to get its basic wireless act together. It needs to get its instruments out there, along with gateways to collect the signals and deliver them to the host. This is where Johannes Bleuel says his company, E-Senza, stands out: while Emerson and Siemens make the HART instruments, E-Senza has the know-how to make the gateways and access points.
‘Our customers do not know the direction the market will go,’ he admits, but many of them are now asking for a gateway hard wired to a dozen access points—maybe even more. ‘Wireless will be only for the last metre,’ he says. E-Senza is developing gateways for Ethernet and, not surprisingly in Europe, Profibus. The company is putting its priority on developing gateways, but selling adapters will come later. He sees a basic adapter selling for about 200 euros. He is working on another adapter, one for hazardous locations, that could connect to 32 instruments simultaneously.
After the first products hit the market, retrofitting won’t be far behind, thinks Thomas Holmes of MACTek Corp. ‘There will be high demand for WirelessHART Adapters especially during the introduction phase of WirelessHART as users retrofit the installed base of wired HART field devices,’ he says, and has already prepared one and two-port adapters. The 1-port screws into an available spare port of the HART field device, while the 2-port is designed to be installed anywhere in the current loop, thanks to MACTek’s PassThrough™ technology.
‘There will be high demand for WirelessHART adapters especially during the introduction phase of WirelessHART as users retrofit the installed base of wired HART field devices,’ says Mr. Holmes. ‘The adapter will retrofit wired HART field devices for 24/7 communication with control, engineering, and asset management systems using Wireless HART. The adaptor also provides users with a low risk means of testing and evaluating Wireless HART in their plants.’
(see diagram in the middle of the main article)
WirelessHART uses 10 msec time division slots to make optimum use of radio time and bandwidth. When a device joins the network, the network manager, which is a part of the gateway, checks its refresh rate parameter and uses this to allocate time slots to ensure the new device has sufficient access to the network to transmit its data. But this is not the end of the story. ‘Smart reporting’ can alter the time slot allocation based upon process conditions. In this example, a process variable is normally refreshed at 20seconds intervals. However it has passed through an alarm condition and the network manager now sees that it increases the reporting frequency to every 10 seconds.
WirelessHART SYSTEM ARCHITECTURE
(see diagram near the end of the main article)
In WirelessHART, each field instrument is a node in the mesh network. It consists of a battery-powered IEEE 802.15.4 radio that operates at 2.4 GHz and continuously hops between 16 channels. (Some of the nodes could be wireless adapters added to older instruments, and some may be repeaters—devices placed in strategic locations to extend the network or route messages around obstacles but without process connections of their own.) The mesh is made more secure as more devices are added to it, because they provide a greater number of radio signal pathways. The access points pick up the transmissions from nearby instrument nodes; multiple access points provide greater path diversity and hence redundancy.
The gateway it a key element in the network. Its equivalent in the wired system is the junction box between the wired HART instruments and the marshalling cabinet. The gateway provides the connection to the host, which may be through Modbus, Profibus, or Ethernet. The host-gateway connection is not specified in WirelessHART and could be copper, fibreoptic or WiFi. The gateway is the ‘remote I/O’ of the system and the access points are the ‘I/O modules.’
In the initial wireless implementations, the gateway will also consist of the network manager and the security manager. The network manager builds and maintains the mesh network, identifies the best paths, and manages the distribution of the 10 msec time slots. Slot access depends on the process value refresh rate and other access requests, such as alarm reporting and configuration changes. The security manager manages and distributes the security encryption keys, and holds the list of authorised devices to join the network.
Contact Details and Archive...
Most Viewed Articles...